Introduction to Probability
and Statistics Using R
G. Jay Kerns
First Edition
BsUR: Introduction to Probability and Statistics Using R
Copyright © 2010 G. Jay Kerns
ISBN: 978-0-557-24979-4
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU
Free Documentation License, Version 1 .3 or any later version published by the Free Software Foun-
dation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the
license is included in the section entitled “GNU Free Documentation License”.
Date: March 24, 2011
Contents
Preface vii
List of Figures xiii
List of Tables xv
1 An Introduction to Probability and Statistics 1
1 . 1 Probability 1
1.2 Statistics 1
Chapter Exercises 2
2 An Introduction to R 3
2.1 Downloading and Installing R 3
2.2 Communicating with R 4
2.3 Basic R Operations and Concepts 6
2.4 Getting Help 12
2.5 External Resources 14
2.6 Other Tips 14
Chapter Exercises 16
3 Data Description 17
3.1 Types of Data 17
3.2 Features of Data Distributions 33
3.3 Descriptive Statistics 35
3.4 Exploratory Data Analysis 41
3.5 Multivariate Data and Data Frames 45
3.6 Comparing Populations 47
Chapter Exercises 55
4 Probability 67
4.1 Sample Spaces 67
4.2 Events 73
4.3 Model Assignment 78
4.4 Properties of Probability 83
4.5 Counting Methods 87
4.6 Conditional Probability 93
iii
CONTENTS
iv
4.7 Independent Events 99
4.8 Bayes’ Rule 102
4.9 Random Variables 106
Chapter Exercises 109
5 Discrete Distributions 111
5.1 Discrete Random Variables 112
5.2 The Discrete Uniform Distribution 114
5.3 The Binomial Distribution 116
5.4 Expectation and Moment Generating Functions 122
5.5 The Empirical Distribution 125
5.6 Other Discrete Distributions 128
5.7 Functions of Discrete Random Variables 136
Chapter Exercises 138
6 Continuous Distributions 143
6.1 Continuous Random Variables 143
6.2 The Continuous Uniform Distribution 148
6.3 The Normal Distribution 149
6.4 Functions of Continuous Random Variables 153
6.5 Other Continuous Distributions 157
Chapter Exercises 164
7 Multivariate Distributions 165
7.1 Joint and Marginal Probability Distributions 166
7.2 Joint and Marginal Expectation 172
7.3 Conditional Distributions 174
7.4 Independent Random Variables 176
7.5 Exchangeable Random Variables 178
7.6 The Bivariate Normal Distribution 179
7.7 Bivariate Transformations of Random Variables 181
7.8 Remarks for the Multivariate Case 184
7.9 The Multinomial Distribution 186
Chapter Exercises 190
8 Sampling Distributions 191
8.1 Simple Random Samples 192
8.2 Sampling from a Normal Distribution 193
8.3 The Central Limit Theorem 196
8.4 Sampling Distributions of Two-Sample Statistics 197
8.5 Simulated Sampling Distributions 200
Chapter Exercises 203
9 Estimation 205
9.1 Point Estimation 205
9.2 Confidence Intervals for Means 214
9.3 Confidence Intervals for Differences of Means 221
CONTENTS
v
9.4 Confidence Intervals for Proportions 223
9.5 Confidence Intervals for Variances 225
9.6 Fitting Distributions 225
9.7 Sample Size and Margin of Error 225
9.8 Other Topics 227
Chapter Exercises 228
10 Hypothesis Testing 229
10.1 Introduction 229
10.2 Tests for Proportions 230
10.3 One Sample Tests for Means and Variances 235
10.4 Two-Sample Tests for Means and Variances 239
10.5 Other Hypothesis Tests 241
10.6 Analysis of Variance 241
10.7 Sample Size and Power 243
Chapter Exercises 248
11 Simple Linear Regression 249
11.1 Basic Philosophy 249
11.2 Estimation 253
11.3 Model Utility and Inference 262
11.4 Residual Analysis 267
11.5 Other Diagnostic Tools 275
Chapter Exercises 283
12 Multiple Linear Regression 285
12.1 The Multiple Linear Regression Model 285
12.2 Estimation and Prediction 288
12.3 Model Utility and Inference 296
12.4 Polynomial Regression 299
12.5 Interaction 304
12.6 Qualitative Explanatory Variables 307
12.7 Partial F Statistic 310
12.8 Residual Analysis and Diagnostic Tools 312
12.9 Additional Topics 313
Chapter Exercises 317
13 Resampling Methods 319
13.1 Introduction 319
13.2 Bootstrap Standard Errors 321
13.3 Bootstrap Confidence Intervals 326
13.4 Resampling in Hypothesis Tests 328
Chapter Exercises 332
14 Categorical Data Analysis 333
15 Nonparametric Statistics 335
vi CONTENTS
16 Time Series 337
A R Session Information 339
B GNU Free Documentation License 341
C History 349
D Data 351
D.l Data Structures 351
D.2 Importing Data 356
D.3 Creating New Data Sets 357
D.4 Editing Data 357
D.5 Exporting Data 359
D. 6 Reshaping Data 359
E Mathematical Machinery 361
E. 1 Set Algebra 361
E.2 Differential and Integral Calculus 362
E.3 Sequences and Series 365
E.4 The Gamma Function 368
E.5 Linear Algebra 368
E. 6 Multivariable Calculus 369
F Writing Reports with R 373
F. l What to Write 373
F.2 How to Write It with R 374
F.3 Formatting Tables 377
F. 4 Other Formats 377
G Instructions for Instructors 379
G. l Generating This Document 380
G.2 How to Use This Document 381
G.3 Ancillary Materials 381
G.4 Modifying This Document 382
H RcradrTestDrive Story 383
Bibliography 389
Index
395
Preface
This book was expanded from lecture materials I use in a one semester upper-division undergradu-
ate course entitled Probability and Statistics at Youngstown State University. Those lecture mate-
rials, in turn, were based on notes that I transcribed as a graduate student at Bowling Green State
University. The course for which the materials were written is 50-50 Probability and Statistics, and
the attendees include mathematics, engineering, and computer science majors (among others). The
catalog prerequisites for the course are a full year of calculus.
The book can be subdivided into three basic parts. The first part includes the introductions and
elementary descriptive statistics', I want the students to be knee-deep in data right out of the gate.
The second part is the study of probability, which begins at the basics of sets and the equally likely
model, journeys past discrete/continuous random variables, and continues through to multivariate
distributions. The chapter on sampling distributions paves the way to the third part, which is in-
ferential statistics. This last part includes point and interval estimation, hypothesis testing, and
finishes with introductions to selected topics in applied statistics.
I usually only have time in one semester to cover a small subset of this book. I cover the material
in Chapter 2 in a class period that is supplemented by a take-home assignment for the students. I
spend a lot of time on Data Description, Probability, Discrete, and Continuous Distributions. I
mention selected facts from Multivariate Distributions in passing, and discuss the meaty parts of
Sampling Distributions before moving right along to Estimation (which is another chapter 1 dwell
on considerably). Hypothesis Testing goes faster after all of the previous work, and by that time
the end of the semester is in sight. 1 normally choose one or two final chapters (sometimes three)
from the remaining to survey, and regret at the end that I did not have the chance to cover more.
In an attempt to be correct 1 have included material in this book which I would normally not
mention during the course of a standard lecture. For instance, I normally do not highlight the
intricacies of measure theory or integrability conditions when speaking to the class. Moreover, I
often stray from the matrix approach to multiple linear regression because many of my students
have not yet been formally trained in linear algebra. That being said, it is important to me for
the students to hold something in their hands which acknowledges the world of mathematics and
statistics beyond the classroom, and which may be useful to them for many semesters to come. It
also mirrors my own experience as a student.
The vision for this document is a more or less self contained, essentially complete, correct,
introductory textbook. There should be plenty of exercises for the student, with full solutions for
some, and no solutions for others (so that the instructor may assign them for grading). By Sweave’s
dynamic nature it is possible to write randomly generated exercises and I had planned to implement
this idea already throughout the book. Alas, there are only 24 hours in a day. Look for more in
future editions.
CONTENTS
viii
Seasoned readers will be able to detect my origins: Probability and Statistical Inference by
Hogg and Tanis [44], Statistical Inference by Casella and Berger [13], and Theory of Point Estima-
tion/Testing Statistical Hypotheses by Lehmann [59, 58], I highly recommend each of those books
to every reader of this one. Some R books with “introductory” in the title that I recommend are
Introductory Statistics with R by Dalgaard [19] and Using R for Introductory Statistics by Verzani
[87], Surely there are many, many other good introductory books about R, but frankly, I have tried
to steer clear of them for the past year or so to avoid any undue influence on my own writing.
I would like to make special mention of two other books: Introduction to Statistical Thought
by Michael Lavine [56] and Introduction to Probability by Grinstead and Snell [37]. Both of these
books are free and are what ultimately convinced me to release HsUR under a free license, too.
Please bear in mind that the title of this book is “Introduction to Probability and Statistics
Using R”, and not “Introduction to R Using Probability and Statistics”, nor even “Introduction
to Probability and Statistics and R Using Words”. The people at the party are Probability and
Statistics; the handshake is R. There are several important topics about R which some individuals
will feel are underdeveloped, glossed over, or wantonly omitted. Some will feel the same way
about the probabilistic and/or statistical content. Still others will just want to learn R and skip all
of the mathematics.
Despite any misgivings: here it is, warts and all. I humbly invite said individuals to take this
book, with the GNU Free Documentation License (GNU-FDL) in hand, and make it better. In that
spirit there are at least a few ways in my view in which this book could be improved.
Better data. The data analyzed in this book are almost entirely from the datasets package in
base R, and here is why:
1. I made a conscious effort to minimize dependence on contributed packages,
2. The data are instantly available, already in the correct format, so we need not take time
to manage them, and
3. The data are real.
I made no attempt to choose data sets that would be interesting to the students; rather, data
were chosen for their potential to convey a statistical point. Many of the data sets are decades
old or more (for instance, the data used to introduce simple linear regression are the speeds
and stopping distances of cars in the 1920’s).
In a perfect world with infinite time I would research and contribute recent, real data in a
context crafted to engage the students in every example. One day I hope to stumble over said
time. In the meantime, I will add new data sets incrementally as time permits.
More proofs. I would like to include more proofs for the sake of completeness (I understand that
some people would not consider more proofs to be improvement). Many proofs have been
skipped entirely, and I am not aware of any rhyme or reason to the current omissions. I will
add more when I get a chance.
More and better graphics: I have not used the ggplot2 package [90] because I do not know how
to use it yet. It is on my to-do list.
More and better exercises: There are only a few exercises in the first edition simply because I
have not had time to write more. I have toyed with the exams package [38] and I believe that
CONTENTS
IX
it is a right way to move forward. As I learn more about what the package can do I would
like to incorporate it into later editions of this book.
About This Document
IPsUR contains many interrelated parts: the Document , the Program , the Package , and the An-
cillaries. In short, the Document is what you are reading right now. The Program provides an
efficient means to modify the Document. The Package is an R package that houses the Program
and the Document. Finally, the Ancillaries are extra materials that reside in the Package and were
produced by the Program to supplement use of the Document. We briefly describe each of them in
turn.
The Document
The Document is that which you are reading right now - UsUR’s raison d’etre. There are trans-
parent copies (nonproprietary text files) and opaque copies (everything else). See the GNU-FDL in
Appendix B for more precise language and details.
IPSUR . tex is a transparent copy of the Document to be typeset with a UTpX distribution such as
MikTgX or TgX Live. Any reader is free to modify the Document and release the modified
version in accordance with the provisions of the GNU-FDL. Note that this file cannot be
used to generate a randomized copy of the Document. Indeed, in its released form it is
only capable of typesetting the exact version of UsUR which you are currently reading.
Furthermore, the . tex file is unable to generate any of the ancillary materials.
IPSUR-xxx.eps, IPSUR-xxx.pdf are the image files for every graph in the Document. These
are needed when typesetting with ETgX.
IPSUR.pdf is an opaque copy of the Document. This is the file that instructors would likely want
to distribute to students.
IPSUR. dvi is another opaque copy of the Document in a different file format.
The Program
The Program includes IPSUR. lyx and its nephew IPSUR. Rnw; the purpose of each is to give
individuals a way to quickly customize the Document for their particular purpose(s).
IPSUR. lyx is the source LyX file for the Program, released under the GNU General Public Li-
cense (GNU GPL) Version 3. This file is opened, modified, and compiled with LyX, a
sophisticated open-source document processor, and may be used (together with Sweave) to
generate a randomized, modified copy of the Document with brand new data sets for some of
the exercises and the solution manuals (in the Second Edition). Additionally, LyX can easily
activate/de activate entire blocks of the document, e.g. the proofs of the theorems, the student
solutions to the exercises, or the instructor answers to the problems, so that the new author
may choose which sections (s)he would like to include in the final Document (again. Second
Edition). The IPSUR. lyx file is all that a person needs (in addition to a properly configured
X
CONTENTS
system - see Appendix G) to generate/compile/export to all of the other formats described
above and below, which includes the ancillary materials IPSUR . Rdata and IPSUR . R.
IPSUR.Rnw is another form of the source code for the Program, also released under the GNU GPL
Version 3. It was produced by exporting IPSUR. lyx into R/Sweave format ( . Rnw). This file
may be processed with Sweave to generate a randomized copy of IPSUR.tex - a transparent
copy of the Document - together with the ancillary materials IPSUR. Rdata and IPSUR. R.
Please note, however, that IPSUR . Rnw is just a simple text file which does not support many
of the extra features that LyX offers such as WYSIWYM editing, instantly (de)activating
branches of the manuscript, and more.
The Package
There is a contributed package on CRAN, called IPSUR. The package affords many advantages, one
being that it houses the Document in an easy-to-access medium. Indeed, a student can have the
Document at his/her fingertips with only three commands:
> install .packages ("IPSUR")
> library (.IPSUR)
> read (IPSUR)
Another advantage goes hand in hand with the Program’s license; since II$UR is free, the source
code must be freely available to anyone that wants it. A package hosted on CRAN allows the author
to obey the license by default.
A much more important advantage is that the excellent facilities at R-Forge are building and
checking the package daily against patched and development versions of the absolute latest pre-
release of R. If any problems surface then I will know about it within 24 hours.
And finally, suppose there is some sort of problem. The package structure makes it incredibly
easy for me to distribute bug-fixes and corrected typographical errors. As an author I can make my
corrections, upload them to the repository at R-Forge, and they will be reflected worldwide within
hours. We aren’t in Kansas anymore, Toto.
Ancillary Materials
These are extra materials that accompany I^UR. They reside in the /etc subdirectory of the
package source.
IPSUR. R is the exported R code from IPSUR.Rnw. With this script, literally every R command
from the entirety of I^UR can be resubmitted at the command line.
Notation
We use the notation x or stem. leaf notation to denote objects, functions, etc.. The sequence
“Statistics > Summaries > Active Dataset’’ means to click the Statistics menu item, next click the
Summaries submenu item, and finally click Active Dataset.
CONTENTS
xi
Acknowledgements
This book would not have been possible without the firm mathematical and statistical foundation
provided by the professors at Bowling Green State University, including Drs. Gabor Szekely, Craig
Zirbel, Arjun K. Gupta, Hanfeng Chen, True Nguyen, and James Albert. I would also like to thank
Drs. Neal Carothers and Kit Chan.
I would also like to thank my colleagues at Youngstown State University for their support.
In particular, I would like to thank Dr. G. Andy Chang for showing me what it means to be a
statistician.
I would like to thank Richard Heiberger for his insightful comments and improvements to
several points and displays in the manuscript.
Finally, and most importantly, I would like to thank my wife for her patience and understanding
while I worked hours, days, months, and years on a free book. In retrospect, I can’t believe I ever
got away with it.
CONTENTS
List of Figures
3.1.1 Strip charts of the precip, rivers, and discoveries data 20
3.1.2 (Relative) frequency histograms of the precip data 22
3.1.3 More histograms of the precip data 23
3.1.4 Index plots of the LakeHuron data 26
3.1.5 Bar graphs of the state . region data 29
3.1.6 Pareto chart of the state . division data 30
3.1.7 Dot chart of the state . region data 31
3.6.1 Boxplots of weight by feed type in the chickwts data 51
3.6.2 Histograms of age by education level from the infert data 52
3.6.3 An xyplot of Petal .Length versus Petal .Width by Species in the iris
data 53
3.6.4 A coplot of cone versus uptake by Type and Treatment in the C02 data . 54
4.0. 1 Two types of experiments 68
4.5.1 The birthday problem 92
5.3.1 Graph of the binom(size = 3, prob = 1/2) CDF 119
5.3.2 The binom(size = 3, prob = 0.5) distribution from the distr package . . . . 121
5.5.1 The empirical CDF 127
6.5.1 Chi square distribution for various degrees of freedom 159
6.5.2 Plot of the gamma( shape = 13, rate = 1) MGF 163
7.6.1 Graph of a bivariate normal PDF 182
7.9.1 Plot of a multinomial PMF 189
8.2.1 Student’s t distribution for various degrees of freedom 195
8.5.1 Plot of simulated IQRs 201
8.5.2 Plot of simulated MADs 202
9.1.1 Capture-recapture experiment 207
9.1.2 Assorted likelihood functions for fishing, part two 209
9.1.3 Species maximum likelihood 210
9.2.1 Simulated confidence intervals 216
9.2.2 Confidence interval plot for the PlantGrowth data 219
Hypothesis test plot based on normal . and . t . di st from the HH package . . . 236
xiii
10.2.1
XIV
LIST OF FIGURES
10.3.1 Hypothesis test plot based on normal . and . t . di st from the HH package . . . 238
10.6.1 Between group versus within group variation 244
10.6.2 Between group versus within group variation 245
10.6.3 Some F plots from the HH package 246
10.7.1 Plot of significance level and power 247
11.1.1 Philosophical foundations of SLR 251
1 1.1.2 Scatterplot of dist versus speed for the cars data 252
1 1.2. 1 Scatterplot with added regression line for the cars data 255
11.2.2 Scatterplot with confidence/prediction bands for the cars data 263
1 1.4. 1 Normal q-q plot of the residuals for the cars data 269
11.4.2 Plot of standardized residuals against the fitted values for the cars data . . . . 271
11.4.3 Plot of the residuals versus the fitted values for the cars data 273
11.5.1 Cook’s distances for the cars data 280
1 1.5.2 Diagnostic plots for the cars data 281
12.1.1 Scatterplot matrix of trees data 287
12.1.2 3D scatterplot with regression plane for the trees data 289
12.4. 1 Scatterplot of Volume versus Girth for the trees data 300
12.4.2 A quadratic model for the trees data 303
12.6.1 A dummy variable model for the trees data 309
13.2.1 Bootstrapping the standard error of the mean, simulated data 322
13.2.2 Bootstrapping the standard error of the median for the rivers data 324
List of Tables
4.1 Sampling k from n objects with urnsamples 89
4.2 Rolling two dice 94
5.1 Correspondence between stats and distr 121
7.1 Maximum U and sum V of a pair of dice rolls (X, Y) 168
7.2 Joint values of U - max(X, Y) and V - X + Y 168
7.3 The joint PMF of (U,V) 168
E. 1 Set operations 361
E.2 Differentiation rules 363
E.3 Some derivatives 363
E.4 Some integrals (constants of integration omitted) 364
xv
LIST OF TABLES
Chapter 1
An Introduction to Probability and
Statistics
This chapter has proved to be the hardest to write, by far. The trouble is that there is so much to
say - and so many people have already said it so much better than I could. When I get something I
like I will release it here.
In the meantime, there is a lot of information already available to a person with an Internet
connection. I recommend to start at Wikipedia, which is not a flawless resource but it has the main
ideas with links to reputable sources.
In my lectures I usually tell stories about Fisher, Galton, Gauss, Laplace, Quetelet, and the
Chevalier de Mere.
1.1 Probability
The common folklore is that probability has been around for millennia but did not gain the attention
of mathematicians until approximately 1654 when the Chevalier de Mere had a question regarding
the fair division of a game’s payoff to the two players, if the game had to end prematurely.
1.2 Statistics
Statistics concerns data; their collection, analysis, and interpretation. In this book we distinguish
between two types of statistics: descriptive and inferential.
Descriptive statistics concerns the summarization of data. We have a data set and we would like
to describe the data set in multiple ways. Usually this entails calculating numbers from the data,
called descriptive measures, such as percentages, sums, averages, and so forth.
Inferential statistics does more. There is an inference associated with the data set, a conclusion
drawn about the population from which the data originated.
I would like to mention that there are two schools of thought of statistics: frequentist and
bayesian. The difference between the schools is related to how the two groups interpret the under-
lying probability (see Section 4.3). The frequentist school gained a lot of ground among statisti-
cians due in large part to the work of Fisher, Neyman, and Pearson in the early twentieth century.
1
2
CHAPTER 1 . AN INTRODUCTION TO PROBABILITY AND STATISTICS
That dominance lasted until inexpensive computing power became widely available; nowadays the
bayesian school is garnering more attention and at an increasing rate.
This book is devoted mostly to the frequentist viewpoint because that is how I was trained, with
the conspicuous exception of Sections 4.8 and 7.3. I plan to add more bayesian material in later
editions of this book.
Chapter Exercises
Chapter 2
An Introduction to R
2.1 Downloading and Installing R
The instructions for obtaining R largely depend on the user’s hardware and operating system. The R
Project has written an R Installation and Administration manual with complete, precise instructions
about what to do, together with all sorts of additional information. The following is just a primer
to get a person started.
2.1.1 Installing R
Visit one of the links below to download the latest version of R for your operating system:
Microsoft Windows: http : //cran . r-pro j ect . org/bin/windows/base/
MacOS: http : //cran . r-pro j ect . org/bin/macosx/
Linux: http : //cran . r-pro j ect . org/bin/linux/
On Microsoft Windows, click the R-x.y.z.exe installer to start installation. When it asks for
"Customized startup options", specify Yes. In the next window, be sure to select the SDI (single
document interface) option; this is useful later when we discuss three dimensional plots with the
rgl package [1].
Installing R on a USB drive (Windows) With this option you can use R portably and without
administrative privileges. There is an entry in the R for Windows FAQ about this. Here is the
procedure I use:
1 . Download the Windows installer above and start installation as usual. When it asks where to
install, navigate to the top-level directory of the USB drive instead of the default C drive.
2. When it asks whether to modify the Windows registry, uncheck the box; we do NOT want to
tamper with the registry.
3. After installation, change the name of the folder from R-x . y . z to just plain R. (Even quicker:
do this in step 1.)
3
4
CHAPTER 2. AN INTRODUCTION TO R
4. Download the following shortcut to the top-level directory of the USB drive, right beside the
R folder, not inside the folder.
http : //ipsur . r- forge . r-pro j ect . org/book/download/R . exe
Use the downloaded shortcut to run R.
Steps 3 and 4 are not required but save you the trouble of navigating to the R-x . y . z/bin directory
to double-click Rgui . exe every time you want to run the program. It is useless to create your own
shortcut to Rgui . exe. Windows does not allow shortcuts to have relative paths; they always have a
drive letter associated with them. So if you make your own shortcut and plug your USB drive into
some other machine that happens to assign your drive a different letter, then your shortcut will no
longer be pointing to the right place.
2.1.2 Installing and Loading Add-on Packages
There are base packages (which come with R automatically), and contributed packages (which
must be downloaded for installation). For example, on the version of R being used for this docu-
ment the default base packages loaded at startup are
> getOption("defaultPackages")
[1] "datasets" "utils" "grDevices" "graphics" "stats"
"methods"
The base packages are maintained by a select group of volunteers, called “R Core”. In addi-
tion to the base packages, there are literally thousands of additional contributed packages written
by individuals all over the world. These are stored worldwide on mirrors of the Comprehensive
R Archive Network, or CRAN for short. Given an active Internet connection, anybody is free to
download and install these packages and even inspect the source code.
To install a package named foo, open up R and type install .packages("foo"). To in-
stall foo and additionally install all of the other packages on which foo depends, instead type
install .packages("foo" , depends = TRUE).
The general command install . packages () will (on most operating systems) open a window
containing a huge list of available packages; simply choose one or more to install.
No matter how many packages are installed onto the system, each one must first be loaded
for use with the library function. For instance, the foreign package [18] contains all sorts of
functions needed to import data sets into R from other software such as SPSS, SAS, etc.. But none
of those functions will be available until the command library(foreign) is issued.
Type library!) at the command prompt (described below) to see a list of all available pack-
ages in your library.
For complete, precise information regarding installation of R and add-on packages, see the R
Installation and Administration manual, http : //cran . r-proj ect . org/manuals . html.
2.2 Communicating with R
One line at a time This is the most basic method and is the first one that beginners will use.
2.2. COMMUNICATING WITH R
5
RGui (Microsoft® Windows)
Terminal
Emacs/ESS, XEmacs
JGR
Multiple lines at a time For longer programs (called scripts) there is too much code to write
all at once at the command prompt. Furthermore, for longer scripts it is convenient to be able
to only modify a certain piece of the script and run it again in R. Programs called script editors
are specially designed to aid the communication and code writing process. They have all sorts of
helpful features including R syntax highlighting, automatic code completion, delimiter matching,
and dynamic help on the R functions as they are being written. Even more, they often have all of
the text editing features of programs like Microsoft® Word. Lastly, most script editors are fully
customizable in the sense that the user can customize the appearance of the interface to choose
what colors to display, when to display them, and how to display them.
R Editor (Windows): In Microsoft® Windows, RGui has its own built-in script editor, called R
Editor. From the console window, select File > New Script. A script window opens, and the
lines of code can be written in the window. When satisfied with the code, the user highlights
all of the commands and presses Ctrl+R. The commands are automatically run at once in R
and the output is shown. To save the script for later, click File > Save as... in R Editor. The
script can be reopened later with File > Open Script... in RGui. Note that R Editor does not
have the fancy syntax highlighting that the others do.
RWinEdt: This option is coordinated with WinEdt for LTpX and has additional features such as
code highlighting, remote sourcing, and a ton of other things. However, one first needs to
download and install a shareware version of another program, WinEdt, which is only free for
a while - pop-up windows will eventually appear that ask for a registration code. RWinEdt
is nevertheless a very fine choice if you already own WinEdt or are planning to purchase it
in the near future.
Tinn-R/Sciviews-K: This one is completely free and has all of the above mentioned options and
more. It is simple enough to use that the user can virtually begin working with it immediately
after installation. But Tinn-R proper is only available for Microsoft® Windows operating
systems. If you are on MacOS or Linux, a comparable alternative is Sci- Views - Komodo
Edit.
Emacs/ESS: Emacs is an all purpose text editor. It can do absolutely anything with respect to
modifying, searching, editing, and manipulating, text. And if Emacs can’t do it, then you
can write a program that extends Emacs to do it. Once such extension is called ESS, which
stands for Emacs .S'peaks Statistics. With ESS a person can speak to R, do all of the tricks that
the other script editors offer, and much, much, more. Please see the following for installation
details, documentation, reference cards, and a whole lot more:
http://ess.r-project.org
6
CHAPTER 2. AN INTRODUCTION TO R
Fair warning: if you want to try Emacs and if you grew up with Microsoft® Windows or
Macintosh, then you are going to need to relearn everything you thought you knew about computers
your whole life. (Or, since Emacs is completely customizable, you can reconfigure Emacs to behave
the way you want.) I have personally experienced this transformation and I will never go back.
JGR (read “Jaguar”): This one has the bells and whistles of RGui plus it is based on Java, so
it works on multiple operating systems. It has its own script editor like R Editor but with
additional features such as syntax highlighting and code-completion. If you do not use
Microsoft® Windows (or even if you do) you definitely want to check out this one.
Kate, Bluefish, etc. There are literally dozens of other text editors available, many of them free,
and each has its own (dis)advantages. I only have mentioned the ones with which I have
had substantial personal experience and have enjoyed at some point. Play around, and let me
know what you find.
Graphical User Interfaces (GUIs) By the word “GUI” I mean an interface in which the user
communicates with R by way of points-and-clicks in a menu of some sort. Again, there are many,
many options and I only mention ones that I have used and enjoyed. Some of the other more
popular script editors can be downloaded from the R-Project website at http : //www . sciviews .
org/_rgui/. On the left side of the screen (under Projects) there are several choices available.
R Commander provides a point-and-click interface to many basic statistical tasks. It is called the
“Commander” because every time one makes a selection from the menus, the code corre-
sponding to the task is listed in the output window. One can take this code, copy-and-paste
it to a text file, then re-run it again at a later time without the R Commander’s assistance.
It is well suited for the introductory level. Rcmdr also allows for user-contributed “Plugins”
which are separate packages on CRAN that add extra functionality to the Rcmdr package. The
plugins are typically named with the prefix RcmdrPlugin to make them easy to identify in
the CRAN package list. One such plugin is the
RcmdrPlugin. IPSUR package which accompanies this text.
Poor Man’s GUI is an alternative to the Rcmdr which is based on GTk instead of Tcl/Tk. It has
been a while since I used it but I remember liking it very much when I did. One thing
that stood out was that the user could drag-and-drop data sets for plots. See here for more
information: http : //wiener . math . csi . cuny . edu/pmg/.
Rattle is a data mining toolkit which was designed to manage/analyze very large data sets, but
it provides enough other general functionality to merit mention here. See [91] for more
information.
Deducer is relatively new and shows promise from what I have seen, but I have not actually used
it in the classroom yet.
2.3 Basic R Operations and Concepts
The R developers have written an introductory document entitled “An Introduction to R”. There is
a sample session included which shows what basic interaction with R looks like. I recommend that
2.3. BASIC R OPERATIONS AND CONCEPTS
1
all new users of R read that document, but bear in mind that there are concepts mentioned which
will be unfamiliar to the beginner.
Below are some of the most basic operations that can be done with R. Almost every book about
R begins with a section like the one below; look around to see all sorts of things that can be done
at this most basic level.
2.3.1 Arithmetic
>2 + 3
# add
[1] 5
>4*5/6
# multiply and divide
[1] 3.333333
> 7A8
# 7 to the 8th power
[1] 5764801
Notice the comment character #. Anything
that 20/6 is a repeating decimal, but the above
number of digits displayed with options:
typed after a # symbol is ignored by R. We know
example shows only 7 digits. We can change the
> options(digits = 16)
> 161/ 3
[1] 3 . BB33BBB33BB333B
> sqrt(2)
[1] 1 .414213562373095
> exp(l)
[1] 2 . 7 182 8 182 84 5904 5
> Pi
[1] 3.141592653589793
> options(digits = 7)
# see more digits
# square root
# Euler's constant, e
# back to default
Note that it is possible to set digits up to 22, but setting them over 16 is not recommended
(the extra significant digits are not necessarily reliable). Above notice the sqrt function for square
roots and the exp function for powers of e, Euler’s number.
2.3.2 Assignment, Object names, and Data types
It is often convenient to assign numbers and values to variables (objects) to be used later. The
proper way to assign values to a variable is with the <- operator (with a space on either side). The
= symbol works too, but it is recommended by the R masters to reserve = for specifying arguments
to functions (discussed later). In this book we will follow their advice and use <- for assignment.
Once a variable is assigned, its value can be printed by simply entering the variable name by itself.
8
CHAPTER 2. AN INTRODUCTION TO R
> x <- 7*41/pi # don't see the calculated value
> x # take a look
[1] 91.35494
When choosing a variable name you can use letters, numbers, dots or underscore
characters. You cannot use mathematical operators, and a leading dot may not be followed by
a number. Examples of valid names are: x, xl, y. value, and y_hat. (More precisely, the set
of allowable characters in object names depends on one’s particular system and locale; see An
Introduction to R for more discussion on this.)
Objects can be of many types, modes, and classes. At this level, it is not necessary to investigate
all of the intricacies of the respective types, but there are some with which you need to become
familiar:
integer: the values 0, ±1, ±2, . . . ; these are represented exactly by R.
double: real numbers (rational and irrational); these numbers are not represented exactly (save
integers or fractions with a denominator that is a power of 2, see [85]).
character: elements that are wrapped with pairs of " or ' ;
logical: includes TRUE, FALSE, and NA (which are reserved words); the NA stands for “not avail-
able”, i.e., a missing value.
You can determine an object’s type with the typeof function. In addition to the above, there is the
complex data type:
> sqrt(-l) # isn't defined
[1] NaN
> sqrt(-l+Qi) # is defined
[1] ®+li
> sqrt (as. complex (-1)) # same thing
[1] ®+li
> (0 + li)A2 # should be -1
[1] -l+®i
> typeof ((Q + li)A2)
[1] "complex"
Note that you can just type (li) A2 to get the same answer. The NaN stands for “not a number”;
it is represented internally as double.
2.3.3 Vectors
All of this time we have been manipulating vectors of length 1 . Now let us move to vectors with
multiple entries.
2.3. BASIC R OPERATIONS AND CONCEPTS
9
Entering data vectors
1. c: If you would like to enter the data 74,31,95,61,76,34,23,54,96 into R, you may
create a data vector with the c function (which is short for concatenate).
> x <- c(74, 31, 95, 61, 76, 34, 23, 54, 96)
> x
[1] 74 31 95 61 76 34 23 54 96
The elements of a vector are usually coerced by R to the the most general type of any of the
elements, so if you do c(l , "2") then the result will be c("l" , "2").
2. scan: This method is useful when the data are stored somewhere else. For instance, you
may type x <- scan() at the command prompt and R will display 1 : to indicate that it is
waiting for the first data value. Type a value and press Enter, at which point R will display
2 : , and so forth. Note that entering an empty line stops the scan. This method is especially
handy when you have a column of values, say, stored in a text file or spreadsheet. You may
copy and paste them all at the 1 : prompt, and R will store all of the values instantly in the
vector x.
3. repeated data; regular patterns: the seq function will generate all sorts of sequences of num-
bers. It has the arguments from, to, by, and length. out which can be set in concert with
one another. We will do a couple of examples to show you how it works.
> seq(from = 1, to - 5)
[1] 1 2 3 4 5
> seq(from =2, by = -0.1, length. out = 4)
[1] 2.® 1.9 1.8 1.7
Note that we can get the first line much quicker with the colon operator :
>1:5
[1] 1 2 3 4 5
The vector LETTERS has the 26 letters of the English alphabet in uppercase and letters has
all of them in lowercase.
Indexing data vectors Sometimes we do not want the whole vector, but just a piece of it. We
can access the intermediate parts with the [] operator. Observe (with x defined above)
> x[l]
[1] 74
> x[2: 4]
10
CHAPTER 2. AN INTRODUCTION TO R
[1] 31 95 61
> x[c(l, 3, 4, 8) ]
[1] 74 95 61 54
> x[-c(l , 3, 4, 8) ]
[1] 31 76 34 23 96
Notice that we used the minus sign to specify those elements that we do not want.
> LETTERS [1:5]
[1] "A" "B" "C" "D"
> letters[-(6:24)]
[1] "a" "b" "c" "d"
2.3.4 Functions and Expressions
A function takes arguments as input and returns an object as output. There are functions to do all
sorts of things. We show some examples below.
> x <- 1:5
> sum(x)
[1] 15
> length(x)
[1] 5
> min(x)
[1] 1
> mean(x) # sample mean
[1] 3
> sd(x) # sample standard deviation
[1] 1.581139
It will not be long before the user starts to wonder how a particular function is doing its job, and
since R is open-source, anybody is free to look under the hood of a function to see how things are
calculated. For detailed instructions see the article “Accessing the Sources” by Uwe Ligges [60].
In short:
1. Type the name of the function without any parentheses or arguments. If you are lucky then
the code for the entire function will be printed, right there looking at you. For instance,
suppose that we would like to see how the intersect function works:
> intersect
2.3. BASIC R OPERATIONS AND CONCEPTS
11
function (x, y)
{
y <- as.vector(y)
unique (y[match(as .vector (x) , y, 0L)])
}
environment : namespace : base>
2. If instead it shows UseMethod ("something") then you will need to choose the class of the
object to be inputted and next look at the method that will be dispatched to the object. For
instance, typing rev says
> rev
function (x)
UseMethod ("rev")
environment : namespace : base>
The output is telling us that there are multiple methods associated with the rev function. To
see what these are, type
> methods (rev)
[1] rev. default rev. dendrogram*
Non-visible functions are asterisked
Now we learn that there are two different rev(x) functions, only one of which being chosen
at each call depending on what x is. There is one for dendrogram objects and a default
method for everything else. Simply type the name to see what each method does. For
example, the default method can be viewed with
> rev. default
function (x)
if (length(x)) x[length(x) : 1L] else x
environment : namespace : base>
3. Some functions are hidden by a namespace (see An Introduction to R [85]), and are not
visible on the first try. For example, if we try to look at the code for wilcox.test (see
Chapter 15) we get the following:
> wilcox.test
function (x, ...)
UseMethod("wilcox . test")
environment : namespace : stats>
12
CHAPTER 2. AN INTRODUCTION TO R
> methods (wilcox. test)
[1] wilcox. test. default* wilcox. test. formula*
Non-visible functions are asterisked
If we were to try wilcox . test . default we would get a “not found” error, because it is
hidden behind the namespace for the package stats (shown in the last line when we tried
wilcox . test). In cases like these we prefix the package name to the front of the func-
tion name with three colons; the command stats : : : wilcox. test . default will show the
source code, omitted here for brevity.
4. Ifit shows . Internal (.something} or . Primitive("somef/i/«g"), then it will be necessary
to download the source code of R (which is not a binary version with an . exe extension)
and search inside the code there. See Ligges [60] for more discussion on this. An example
is exp:
> exp
function (x) .Primitive ("exp")
Be warned that most of the . Internal functions are written in other computer languages
which the beginner may not understand, at least initially.
2.4 Getting Help
When you are using R, it will not take long before you find yourself needing help. Fortunately,
R has extensive help resources and you should immediately become familiar with them. Begin by
clicking Help on Rgui. The following options are available.
• Console: gives useful shortcuts, for instance, Ctrl+L. to clear the R console screen.
• FAQ on R: frequently asked questions concerning general R operation.
• FAQ on R for Windows: frequently asked questions about R, tailored to the Microsoft
Windows operating system.
• Manuals: technical manuals about all features of the R system including installation, the
complete language definition, and add-on packages.
• R functions (text). . . : use this if you know the exact name of the function you want to know
more about, for example, mean or plot. Typing mean in the window is equivalent to typing
help ("mean") at the command line, or more simply, ?mean. Note that this method only
works if the function of interest is contained in a package that is already loaded into the
search path with library.
• HTML Help: use this to browse the manuals with point-and-click links. It also has a Search
Engine & Keywords for searching the help page titles, with point-and-click links for the
search results. This is possibly the best help method for beginners. It can be started from the
command line with the command help . start () .
2.4. GETTING HELP
13
• Search help. . . : use this if you do not know the exact name of the function of interest, or if
the function is in a package that has not been loaded yet. For example, you may enter plo
and a text window will return listing all the help hies with an alias, concept, or title matching
‘plo’ using regular expression matching; it is equivalent to typing help. search("plo")
at the command line. The advantage is that you do not need to know the exact name of
the function; the disadvantage is that you cannot point-and-click the results. Therefore, one
may wish to use the HTML Help search engine instead. An equivalent way is ??plo at the
command line.
• search.r-project.org. . . : this will search for words in help lists and email archives of the R
Project. It can be very useful for finding other questions that other users have asked.
• Apropos. . . : use this for more sophisticated partial name matching of functions. See ?apropos
for details.
On the help pages for a function there are sometimes “Examples” listed at the bottom of the page,
which will work if copy-pasted at the command line (unless marked otherwise). The example
function will run the code automatically, skipping the intermediate step. For instance, we may try
example (mean) to see a few examples of how the mean function works.
2.4.1 R Help Mailing Lists
There are several mailing lists associated with R, and there is a huge community of people that
read and answer questions related to R. See here http : //www. r- project . org/mail .html for
an idea of what is available. Particularly pay attention to the bottom of the page which lists several
special interest groups (SIGs) related to R.
Bear in mind that R is free software, which means that it was written by volunteers, and the
people that frequent the mailing lists are also volunteers who are not paid by customer support fees.
Consequently, if you want to use the mailing lists for free advice then you must adhere to some
basic etiquette, or else you may not get a reply, or even worse, you may receive a reply which is a
bit less cordial than you are used to. Below are a few considerations:
1. Read the FAQ (http://cran.r-project.org/faqs.html). Note that there are different
FAQs for different operating systems. You should read these now, even without a question at
the moment, to learn a lot about the idiosyncrasies of R.
2. Search the archives. Even if your question is not a FAQ, there is a very high likelihood that
your question has been asked before on the mailing list. If you want to know about topic f oo,
then you can do RSiteSearch("foo") to search the mailing list archives (and the online
help) for it.
3. Do a Google search and an RSeek . org search.
If your question is not a FAQ, has not been asked on R-help before, and does not yield to a Google
(or alternative) search, then, and only then, should you even consider writing to R-help. Below are
a few additional considerations.
1. Read the posting guide (http://www.r-project.org/posting-guide.html) before
posting. This will save you a lot of trouble and pain.
14 CHAPTER 2. AN INTRODUCTION TO R
2. Get rid of the command prompts (>) from output. Readers of your message will take the text
from your mail and copy-paste into an R session. If you make the readers’ job easier then it
will increase the likelihood of a response.
3. Questions are often related to a specific data set, and the best way to communicate the data is
with a dump command. For instance, if your question involves data stored in a vector x, you
can type dump("x" , at the command prompt and copy-paste the output into the body of
your email message. Then the reader may easily copy-paste the message from your email
into R and x will be available to him/her.
4. Sometimes the answer the question is related to the operating system used, the attached
packages, or the exact version of R being used. The sessionlnfoO command collects
all of this information to be copy -pasted into an email (and the Posting Guide requests this
information). See Appendix A for an example.
2.5 External Resources
There is a mountain of information on the Internet about R. Below are a few of the important ones.
The R Project for Statistical Computing: (http : //www . r-pro j ect . org/) Go here first.
The Comprehensive R Archive Network: (http://cran.r-project.org/) This is where R
is stored along with thousands of contributed packages. There are also loads of contributed
information (books, tutorials, etc.). There are mirrors all over the world with duplicate infor-
mation.
R-Forge: (http://r-forge.r-project.org/) This is another location where R packages are
stored. Here you can find development code which has not yet been released to CRAN.
R Wiki: (http : //wiki . r- project . org/rwiki/doku .php) There are many tips and tricks listed
here. If you find a trick of your own, login and share it with the world.
Other: the R Graph Gallery (http : //addictedtor . free . fr/graphiques/) and R Graphical
Manual (http://bm2.genes.nig.ac.jp/RGM2/index.php) have literally thousands of
graphs to peruse. RSeek (http://www.rseek.org) is a search engine based on Google
specifically tailored for R queries.
2.6 Other Tips
It is unnecessary to retype commands repeatedly, since R remembers what you have recently en-
tered on the command line. On the Microsoft® Windows RGui, to cycle through the previous
commands just push the j (up arrow) key. On Emacs/ESS the command is M-p (which means hold
down the Alt button and press “p”). More generally, the command historyO will show a whole
list of recently entered commands.
• To find out what all variables are in the current work environment, use the commands
objectsO orlsQ. These list all available objects in the workspace. If you wish to remove
one or more variables, use remove(varl , var2 , var3), or more simply use rm(varl ,
var2 , var 3), and to remove all objects use rm(li st = lsQ).
2.6. OTHER TIPS
15
• Another use of scan is when you have a long list of numbers (separated by spaces or on
different lines) already typed somewhere else, say in a text file. To enter all the data in one
fell swoop, first highlight and copy the list of numbers to the Clipboard with Edit > Copy
(or by right-clicking and selecting Copy). Next type the x <- scanQ command in the R
console, and paste the numbers at the 1 : prompt with Edit > Paste. All of the numbers will
automatically be entered into the vector x.
• The command Ctrl+1 clears the screen in the Microsoft® Windows RGui. The comparable
command for Emacs/ESS is
• Once you use R for awhile there may be some commands that you wish to run automatically
whenever R starts. These commands may be saved in a file called Rprofile . site which
is usually in the etc folder, which lives in the R home directory (which on Microsoft®
Windows usually is C : \Program Files\R). Alternatively, you can make a file . Rprofile
to be stored in the user’s home directory, or anywhere R is invoked. This allows for multiple
configurations for different projects or users. See “Customizing the Environment” of An
Introduction to R for more details.
• When exiting R the user is given the option to “save the workspace”. I recommend that
beginners DO NOT save the workspace when quitting. If Yes is selected, then all of the
objects and data currently in R’s memory is saved in a file located in the working directory
called .RData. This file is then automatically loaded the next time R starts (in which case
R will say [previously saved workspace restored]). This is a valuable feature for
experienced users of R, but 1 find that it causes more trouble than it saves with beginners.
16
CHAPTER 2. AN INTRODUCTION TO R
Chapter Exercises
Chapter 3
Data Description
In this chapter we introduce the different types of data that a statistician is likely to encounter, and in
each subsection we give some examples of how to display the data of that particular type. Once we
see how to display data distributions, we next introduce the basic properties of data distributions.
We qualitatively explore several data sets. Once that we have intuitive properties of data sets,
we next discuss how we may numerically measure and describe those properties with descriptive
statistics.
What do I want them to know?
• different data types, such as quantitative versus qualitative, nominal versus ordinal, and dis-
crete versus continuous
• basic graphical displays for assorted data types, and some of their ( disadvantages
• fundamental properties of data distributions, including center, spread, shape, and crazy ob-
servations
• methods to describe data (visually/numerically) with respect to the properties, and how the
methods differ depending on the data type
• all of the above in the context of grouped data, and in particular, the concept of a factor
3.1 Types of Data
Loosely speaking, a datum is any piece of collected information, and a data set is a collection of
data related to each other in some way. We will categorize data into five types and describe each in
turn:
Quantitative data associated with a measurement of some quantity on an observational unit.
Qualitative data associated with some quality or property of the observational unit.
Logical data to represent true or false and which play an important role later.
17
18
CHAPTER 3. DATA DESCRIPTION
Missing data that should be there but are not, and
Other types everything else under the sun.
In each subsection we look at some examples of the type in question and introduce methods to
display them.
3.1.1 Quantitative data
Quantitative data are any data that measure or are associated with a measurement of the quantity of
something. They invariably assume numerical values. Quantitative data can be further subdivided
into two categories.
• Discrete data take values in a finite or countably infinite set of numbers, that is, all possible
values could (at least in principle) be written down in an ordered list. Examples include:
counts, number of arrivals, or number of successes. They are often represented by integers,
say, 0, 1,2, etc..
• Continuous data take values in an interval of numbers. These are also known as scale data,
interval data, or measurement data. Examples include: height, weight, length, time, etc.
Continuous data are often characterized by fractions or decimals: 3.82, 7.0001, 4 |, etc..
Note that the distinction between discrete and continuous data is not always clear-cut. Sometimes
it is convenient to treat data as if they were continuous, even though strictly speaking they are not
continuous. See the examples.
Example 3.1. Annual Precipitation in US Cities. The vector precip contains average amount
of rainfall (in inches) for each of 70 cities in the United States and Puerto Rico. Let us take a look
at the data:
> str (precip)
Named num [1:70] 67 54
- attr(*, "names")= ch
> precip [1:4]
Mobile Juneau
67.0 54.7
The output shows that precip is a numeric vector which has been named , that is, each value
has a name associated with it (which can be set with the names function). These are quantitative
continuous data.
Example 3.2. Lengths of Major North American Rivers. The U.S. Geological Survey recorded
the lengths (in miles) of several rivers in North America. They are stored in the vector rivers in
the datasets package (which ships with base R). See ?rivers. Let us take a look at the data with
the str function.
.7 7 48.5 14 17.2 20.7 13 43.4 40.2 ...
r [1:70] "Mobile" "Juneau" "Phoenix" "Little Rock"
Phoenix Little Rock
7.0 48.5
> str (rivers)
num [1:141] 735 320 325 392 524 ...
3.1. TYPES OF DATA
19
The output says that rivers is a numeric vector of length 141, and the first few values are 735,
320, 325, etc. These data are definitely quantitative and it appears that the measurements have been
rounded to the nearest mile. Thus, strictly speaking, these are discrete data. But we will find it
convenient later to take data like these to be continuous for some of our statistical procedures.
Example 3.3. Yearly Numbers of Important Discoveries. The vector discoveries contains
numbers of “great” inventions/discoveries in each year from 1860 to 1959, as reported by the 1975
World Almanac. Let us take a look at the data:
> str (discoveries)
Time-Series [1:10®] from 1860 to 1959: 5302032361 ...
> discoveries [1:4]
[1] 5 3 0 2
The output is telling us that discoveries is a time series (see Section 3.1.5 for more) of length
100. The entries are integers, and since they represent counts this is a good example of discrete
quantitative data. We will take a closer look in the following sections.
Displaying Quantitative Data
One of the first things to do when confronted by quantitative data (or any data, for that matter) is
to make some sort of visual display to gain some insight into the data’s structure. There are almost
as many display types from which to choose as there are data sets to plot. We describe some of the
more popular alternatives.
Strip charts (also known as Dot plots) These can be used for discrete or continuous data, and
usually look best when the data set is not too large. Along the horizontal axis is a numerical scale
above which the data values are plotted. We can do it in R with a call to the stripchart function.
There are three available methods.
overplot plots ties covering each other. This method is good to display only the distinct values
assumed by the data set.
jitter adds some noise to the data in the y direction in which case the data values are not covered
up by ties.
stack plots repeated values stacked on top of one another. This method is best used for discrete
data with a lot of ties; if there are no repeats then this method is identical to overplot.
See Figure 3.1.1, which is produced by the following code.
> stripchart (precip, xlab = "rainfall")
> stripchart (rivers, method = "jitter", xlab = "length")
> stripchart (discoveries , method = "stack", xlab = "number")
20
CHAPTER 3. DATA DESCRIPTION
— i — i — i — i — i — i — r
0 1000 2500
length
Figure 3.1.1: Strip charts of the precip, rivers, and discoveries data
The first graph uses the overplot method, the second the jitter method, and the third the stack method.
The leftmost graph is a strip chart of the precip data. The graph shows tightly clustered values
in the middle with some others falling balanced on either side, with perhaps slightly more falling
to the left. Later we will call this a symmetric distribution, see Section 3.2.3. The middle graph is
of the rivers data, a vector of length 141. There are several repeated values in the rivers data, and
if we were to use the overplot method we would lose some of them in the display. This plot shows
a what we will later call a right-skewed shape with perhaps some extreme values on the far right of
the display. The third graph strip charts discoveries data which are literally a textbook example
of a right skewed distribution.
The DOTplot function in the UsingR package [86] is another alternative.
Histogram These are typically used for continuous data. A histogram is constructed by first
deciding on a set of classes, or bins, which partition the real line into a set of boxes into which the
data values fall. Then vertical bars are drawn over the bins with height proportional to the number
3.1. TYPES OF DATA
21
of observations that fell into the bin.
These are one of the most common summary displays, and they are often misidentified as “Bar
Graphs” (see below.) The scale on the y axis can be frequency, percentage, or density (relative
frequency). The term histogram was coined by Karl Pearson in 1891, see [66].
Example 3.4. Annual Precipitation in US Cities. We are going to take another look at the precip
data that we investigated earlier. The strip chart in Figure 3.1.1 suggested a loosely balanced
distribution; let us now look to see what a histogram says.
There are many ways to plot histograms in R, and one of the easiest is with the hist function.
The following code produces the plots in Figure 3.1.2.
> hist (precip, main =
> hist (precip, freq = FALSE, main =
Notice the argument main = which suppresses the main title from being displayed - it
would have said “Histogram of precip” otherwise. The plot on the left is a frequency histogram
(the default), and the plot on the right is a relative frequency histogram (freq = FALSE).
Please be careful regarding the biggest weakness of histograms: the graph obtained strongly
depends on the bins chosen. Choose another set of bins, and you will get a different histogram.
Moreover, there are not any definitive criteria by which bins should be defined; the best choice for
a given data set is the one which illuminates the data set’s underlying structure (if any). Luckily for
us there are algorithms to automatically choose bins that are likely to display well, and more often
than not the default bins do a good job. This is not always the case, however, and a responsible
statistician will investigate many bin choices to test the stability of the display.
Example 3.5. Recall that the strip chart in Figure 3.1.1 suggested a relatively balanced shape to the
precip data distribution. Watch what happens when we change the bins slightly (with the breaks
argument to hist). See Figure 3.1.3 which was produced by the following code.
> hist(precip, breaks = US, main =
> hist(precip, breaks = 2(90, main = "")
The leftmost graph (with breaks = 10) shows that the distribution is not balanced at all. There
are two humps: a big one in the middle and a smaller one to the left. Graphs like this often indicate
some underlying group structure to the data; we could now investigate whether the cities for which
rainfall was measured were similar in some way, with respect to geographic region, for example.
The rightmost graph in Figure 3.1.3 shows what happens when the number of bins is too large:
the histogram is too grainy and hides the rounded appearance of the earlier histograms. If we were
to continue increasing the number of bins we would eventually get all observed bins to have exactly
one element, which is nothing more than a glorified strip chart.
Stemplots (more to be said in Section 3.4) Stemplots have two basic parts: stems and leaves.
The final digit of the data values is taken to be a leaf, and the leading digit(s) is (are) taken to be
stems. We draw a vertical line, and to the left of the line we list the stems. To the right of the line,
we list the leaves beside their corresponding stem. There will typically be several leaves for each
stem, in which case the leaves accumulate to the right. It is sometimes necessary to round the data
values, especially for larger data sets.
Frequency
22
CHAPTER 3. DATA DESCRIPTION
precip precip
Figure 3.1.2: (Relative) frequency histograms of the precip data
3.1. TYPES OF DATA
23
precip
precip
Figure 3.1.3: More histograms of the precip data
24
CHAPTER 3. DATA DESCRIPTION
Example 3.6. UKDriverDeaths is a time series that contains the total car drivers killed or se-
riously injured in Great Britain monthly from Jan 1969 to Dec 1984. See ? UKDriverDeaths.
Compulsory seat belt use was introduced on January 31, 1983. We construct a stem and leaf dia-
gram in R with the stem, leaf function from the aplpack package [92].
> library(aplpack)
> stem. leaf (UKDriverDeaths, depth = FALSE)
1 | 2: represents 12®
leaf unit: 1®
n: 192
1® |
11 I
12 |
13 I
14 |
15 I
16 I
17 I
18 I
19 I
2® |
21 |
22 |
23 |
24 |
HI: 2654
57
136678
123889
Q255666888899
Q®8® 12223444445 5 5 556667788889
®88® 111 11222222344445 5 555 566677779
81222333444445555555678888889
11233344566667799
88811235568
81234455667799
8888113557788899
145599
813467
9
7
The display shows a more or less balanced mound-shaped distribution, with one or maybe two
humps, a big one and a smaller one just to its right. Note that the data have been rounded to the
tens place so that each datum gets only one leaf to the right of the dividing line.
Notice that the depths have been suppressed. To learn more about this option and many others,
see Section 3.4. Unlike a histogram, the original data values may be recovered from the stemplot
display - modulo the rounding - that is, starting from the top and working down we can read off
the data values 1050, 1070, 1110, 1130, etc.
Index plot Done with the plot function. These are good for plotting data which are ordered,
for example, when the data are measured over time. That is, the first observation was measured at
time 1, the second at time 2, etc. It is a two dimensional plot, in which the index (or time) is the
x variable and the measured value is the y variable. There are several plotting methods for index
plots, and we discuss two of them:
spikes: draws a vertical line from the x-axis to the observation height (type = "h").
points: plots a simple point at the observation height (type = "p").
Example 3.7. Level of Lake Huron 1875-1972. Brockwell and Davis [11] give the annual mea-
surements of the level (in feet) of Lake Huron from 1875-1972. The data are stored in the time
series LakeHuron. See ?LakeHuron. Figure 3.1.4 was produced with the following code:
3.1. TYPES OF DATA
25
> plot(LakeHuron, type = "h")
> plot(LakeHuron, type = "p ")
The plots show an overall decreasing trend to the observations, and there appears to be some
seasonal variation that increases over time.
Density estimates
3.1.2 Qualitative Data, Categorical Data, and Factors
Qualitative data are simply any type of data that are not numerical, or do not represent numerical
quantities. Examples of qualitative variables include a subject’s name, gender, race/ethnicity, po-
litical party, socioeconomic status, class rank, driver’s license number, and social security number
(SSN).
Please bear in mind that some data look to be quantitative but are not, because they do not
represent numerical quantities and do not obey mathematical rules. For example, a person’s shoe
size is typically written with numbers: 8, or 9, or 12, or 12 | . Shoe size is not quantitative, however,
because if we take a size 8 and combine with a size 9 we do not get a size 17.
Some qualitative data serve merely to identify the observation (such a subject’s name, driver’s
license number, or SSN). This type of data does not usually play much of a role in statistics. But
other qualitative variables serve to subdivide the data set into categories; we call these factors. In
the above examples, gender, race, political party, and socioeconomic status would be considered
factors (shoe size would be another one). The possible values of a factor are called its levels. For
instance, the factor gender would have two levels, namely, male and female. Socioeconomic status
typically has three levels: high, middle, and low.
Factors may be of two types: nominal and ordinal. Nominal factors have levels that correspond
to names of the categories, with no implied ordering. Examples of nominal factors would be hair
color, gender, race, or political party. There is no natural ordering to “Democrat” and “Republican”;
the categories are just names associated with different groups of people.
In contrast, ordinal factors have some sort of ordered structure to the underlying factor levels.
For instance, socioeconomic status would be an ordinal categorical variable because the levels cor-
respond to ranks associated with income, education, and occupation. Another example of ordinal
categorical data would be class rank.
Factors have special status in R. They are represented internally by numbers, but even when they
are written numerically their values do not convey any numeric meaning or obey any mathematical
rules (that is. Stage III cancer is not Stage I cancer + Stage II cancer).
Example 3.8. The state . abb vector gives the two letter postal abbreviations for all 50 states.
> str (state. abb)
chr [1:50] "AL" "AK" "AZ" "AR" "CA" "CO" "CT" "DE"
These would be ID data. The state . name vector lists all of the complete names and those data
would also be ID.
Example 3.9. U.S. State Facts and Features. The U.S. Department of Commerce of the U.S. Cen-
sus Bureau releases all sorts of information in the Statistical Abstract of the United States , and the
LakeHuron LakeHuron
26
CHAPTER 3. DATA DESCRIPTION
CM
1880 1900 1920 1940 1960
Time
CM
1880 1900 1920 1940 1960
Time
Figure 3.1.4: Index plots of the LakeHuron data
3.1. TYPES OF DATA
21
state . region data lists each of the 50 states and the region to which it belongs, be it Northeast,
South, North Central, or West. See ?state . region.
> str (state . region)
Factor w/ 4 levels "Northeast" , "South" : 2442441222 ...
> state. region[l: 5]
[1] South West West South West
Levels: Northeast South North Central West
The str output shows that state . region is already stored internally as a factor and it lists a
couple of the factor levels. To see all of the levels we printed the first five entries of the vector in
the second line. need to print a piece of the from
Displaying Qualitative Data
Tables One of the best ways to summarize qualitative data is with a table of the data values. We
may count frequencies with the table function or list proportions with the prop . table function
(whose input is a frequency table). In the R Commander you can do it with Statistics > Frequency
Distribution Alternatively, to look at tables for all factors in the Active data set you can do
Statistics > Summaries > Active Dataset.
> Tbl <- table(state. division)
> Tbl # frequencies
state . division
New England Middle Atlantic South Atlantic
6 B 8
East South Central West South Central East North Central
4 4 5
West North Central Mountain Pacific
7 8 5
> Tbl/sum(Tbl) # relative frequencies
state . division
New England Middle Atlantic
0.12 0.06
East South Central West South Central
0.08 0.08
West North Central Mountain
0.14 0.16
South Atlantic
0.16
East North Central
0.10
Pacific
0.10
> prop. table (Tbl)
state . division
New England
0.12
East South Central
0.08
West North Central
0.14
# same thing
Middle Atlantic
0.06
West South Central
0.08
Mountain
0.16
South Atlantic
0.16
East North Central
0.10
Pacific
0.10
28
CHAPTER 3. DATA DESCRIPTION
Bar Graphs A bar graph is the analogue of a histogram for categorical data. A bar is displayed
for each level of a factor, with the heights of the bars proportional to the frequencies of observations
falling in the respective categories. A disadvantage of bar graphs is that the levels are ordered
alphabetically (by default), which may sometimes obscure patterns in the display.
Example 3.10. U.S. State Facts and Features. The state. region data lists each of the 50
states and the region to which it belongs, be it Northeast, South, North Central, or West. See
?state . region. It is already stored internally as a factor. We make a bar graph with the barplot
function:
> barplot (table(state. region) , cex. names = 19.5)
> barplot (prop, table (table (state, region)) , cex. names = 19.5)
See Figure 3.1.5. The display on the left is a frequency bar graph because the y axis shows
counts, while the display on the left is a relative frequency bar graph. The only difference between
the two is the scale. Looking at the graph we see that the majority of the fifty states are in the South,
followed by West, North Central, and finally Northeast. Over 30% of the states are in the South.
Notice the cex. names argument that we used, above. It shrinks the names on the r axis by
50% which makes them easier to read. See ?par for a detailed list of additional plot parameters.
Pareto Diagrams A pareto diagram is a lot like a bar graph except the bars are rearranged such
that they decrease in height going from left to right. The rearrangement is handy because it can
visually reveal structure (if any) in how fast the bars decrease - this is much more difficult when
the bars are jumbled.
Example 3.11. U.S. State Facts and Features. The state . division data record the division
(New England, Middle Atlantic, South Atlantic, East South Central, West South Central, East
North Central, West North Central, Mountain, and Pacific) of the fifty states. We can make a pareto
diagram with either the RcmdrPlugin . IPSUR package or with the pareto . chart function from
the qcc package [77]. See Figure 3.1.6. The code follows.
> library (qcc)
> pareto .chart (table (state .division) , ylab = "Frequency")
Dot Charts These are a lot like a bar graph that has been turned on its side with the bars replaced
by dots on horizontal lines. They do not convey any more (or less) information than the associated
bar graph, but the strength lies in the economy of the display. Dot charts are so compact that it
is easy to graph very complicated multi-variable interactions together in one graph. See Section
3.6. We will give an example here using the same data as above for comparison. The graph was
produced by the following code.
> x <- table (state. region)
> dotchart (as. vector (x) , labels = names(x))
See Figure 3.1.7. Compare it to Figure 3.1.5.
3.1. TYPES OF DATA
29
Northeast South West Northeast South
Figure 3.1.5: Bar graphs of the state. region data
West
The left graph is a frequency barplot made with table and the right is a relative frequency barplot made with
prop, table.
30
CHAPTER 3. DATA DESCRIPTION
Package qcc , version 2.8.1
Type ' citation("qcc") ' for citing this R package in publications
Pareto chart analysis for table (state. division)
Frequency Cum.Freq. Percentage Cum. Percent.
Mountain 8
South Atlantic 8
West North Central 7
New England 6
Pacific 5
East North Central 5
West South Central 4
East South Central 4
Middle Atlantic 3
8
16
16
16
16
32
23
14
46
29
12
58
34
10
68
39
10
78
43
8
86
47
8
94
50
6
100
Pareto Chart for table(state.division)
Figure 3.1.6: Pareto chart of the state . division data
Cumulative Percentage
32
CHAPTER 3. DATA DESCRIPTION
Pie Graphs These can be done with R and the R Commander, but they fallen out of favor in
recent years because researchers have determined that while the human eye is good at judging
linear measures, it is notoriously bad at judging relative areas (such as those displayed by a pie
graph). Pie charts are consequently a very bad way of displaying information. A bar chart or dot
chart is a preferable way of displaying qualitative data. See ?pie for more information.
We are not going to do any examples of a pie graph and discourage their use elsewhere.
3.1.3 Logical Data
There is another type of information recognized by R which does not fall into the above categories.
The value is either TRUE or FALSE (note that equivalently you can use 1 = TRUE, ® = FALSE).
Here is an example of a logical vector:
> x <- 5:9
> y <- (x < 7.3)
> y
[1] TRUE TRUE TRUE FALSE FALSE
Many functions in R have options that the user may or may not want to activate in the function
call. For example, the stem, leaf function has the depths argument which is TRUE by default. We
saw in Section 3.1.1 how to turn the option off, simply enter stem. leaf(x, depths = FALSE)
and they will not be shown on the display.
We can swap TRUE with FALSE with the exclamation point ! .
> !y
[1] FALSE FALSE FALSE TRUE TRUE
3.1.4 Missing Data
Missing data are a persistent and prevalent problem in many statistical analyses, especially those
associated with the social sciences. R reserves the special symbol NA to representing missing data.
Ordinary arithmetic with NA values give NA’s (addition, subtraction, etc.) and applying a func-
tion to a vector that has an NA in it will usually give an NA.
> x <- c(3, 7, NA, 4, 7)
> y <- c(5 , NA, 1, 2, 2)
> x + y
[1] 8 NA NA 6 9
Some functions have a na . rm argument which when TRUE will ignore missing data as if it were
not there (such as mean, var, sd, IQR, mad, . . . ).
> sum(x)
[1] NA
> sum(x, na.rm = TRUE)
3.2. FEATURES OF DATA DISTRIBUTIONS
33
[1] 21
Other functions do not have a na . rm argument and will return NA or an error if the argument
has NAs. In those cases we can find the locations of any NAs with the is .na function and remove
those cases with the [] operator.
> is.na(x)
[1] FALSE FALSE TRUE FALSE FALSE
> z <- x[ ! is .na(x) ]
> sum(z)
[1] 21
The analogue of is . na for rectangular data sets (or data frames) is the complete . cases func-
tion. See Appendix D.4.
3.1.5 Other Data Types
3.2 Features of Data Distributions
Given that the data have been appropriately displayed, the next step is to try to identify salient
features represented in the graph. The acronym to remember is Center, f/n usual features, Spread,
and Shape. (CUSS).
3.2.1 Center
One of the most basic features of a data set is its center. Loosely speaking, the center of a data set
is associated with a number that represents a middle or general tendency of the data. Of course,
there are usually several values that would serve as a center, and our later tasks will be focused on
choosing an appropriate one for the data at hand. Judging from the histogram that we saw in Figure
3.1.3, a measure of center would be about 35.
3.2.2 Spread
The spread of a data set is associated with its variability; data sets with a large spread tend to cover
a large interval of values, while data sets with small spread tend to cluster tightly around a central
value.
3.2.3 Shape
When we speak of the shape of a data set, we are usually referring to the shape exhibited by an
associated graphical display, such as a histogram. The shape can tell us a lot about any underlying
structure to the data, and can help us decide which statistical procedure we should use to analyze
them.
34
CHAPTER 3. DATA DESCRIPTION
Symmetry and Skewness A distribution is said to be right-skewed (or positively skewed ) if the
right tail seems to be stretched from the center. A left-skewed (or negatively skewed ) distribution
is stretched to the left side. A symmetric distribution has a graph that is balanced about its center,
in the sense that half of the graph may be reflected about a central line of symmetry to match the
other half.
We have already encountered skewed distributions: both the discoveries data in Figure 3.1.1
and the precip data in Figure 3.1.3 appear right-skewed. The UKDriverDeaths data in Example
3.6 is relatively symmetric (but note the one extreme value 2654 identified at the bottom of the
stemplot).
Kurtosis Another component to the shape of a distribution is how “peaked” it is. Some distri-
butions tend to have a flat shape with thin tails. These are called platykurtic , and an example of a
platykurtic distribution is the uniform distribution; see Section 6.2. On the other end of the spec-
trum are distributions with a steep peak, or spike, accompanied by heavy tails; these are called
leptokurtic. Examples of leptokurtic distributions are the Laplace distribution and the logistic dis-
tribution. See Section 6.5. In between are distributions (called mesokurtic) with a rounded peak
and moderately sized tails. The standard example of a mesokurtic distribution is the famous bell-
shaped curve, also known as the Gaussian, or normal, distribution, and the binomial distribution
can be mesokurtic for specific choices of p. See Sections 5.3 and 6.3.
3.2.4 Clusters and Gaps
Clusters or gaps are sometimes observed in quantitative data distributions. They indicate clumping
of the data about distinct values, and gaps may exist between clusters. Clusters often suggest an
underlying grouping to the data. For example, take a look at the faithful data which contains
the duration of eruptions and the waiting time between eruptions of the Old Faithful geyser in
Yellowstone National Park. (Do not be frightened by the complicated information at the left of the
display for now; we will learn how to interpret it in Section 3.4).
> library (aplpack)
> stem. leaf (faithful$eruptions)
1 | 2: represents 1.2
leaf unit:
0.1
n: 272
12
s
| 667777777777
51
1.
I 888888888888888888888888888899999999999
71
2*
| 00000000000011111111
87
t
| 2222222222333333
92
f
| 44444
94
s
I 66
97
2.
I 889
98
3*
1 o
102
t
I 3333
108
f
I 445555
118
s
| 6666677777
(16)
3.
| 8888888889999999
3.3. DESCRIPTIVE STATISTICS
35
138
187
78
43
21
4
4*
t
£
s
4.
5*
8Q0QQ888888Q0Q881 1 1111111111111
22222222222233333333333333333
44444444444445555555555555555555555
6666666666677777777777
88888888888899999
0001
There are definitely two clusters of data here; an upper cluster and a lower cluster.
3.2.5 Extreme Observations and other Unusual Features
Extreme observations fall far from the rest of the data. Such observations are troublesome to many
statistical procedures; they cause exaggerated estimates and instability. It is important to identify
extreme observations and examine the source of the data more closely. There are many possible
reasons underlying an extreme observation:
• Maybe the value is a typographical error. Especially with large data sets becoming more
prevalent, many of which being recorded by hand, mistakes are a common problem. After
closer scrutiny, these can often be fixed.
• Maybe the observation was not meant for the study, because it does not belong to the
population of interest. For example, in medical research some subjects may have relevant
complications in their genealogical history that would rule out their participation in the ex-
periment. Or when a manufacturing company investigates the properties of one of its devices,
perhaps a particular product is malfunctioning and is not representative of the majority of the
items.
• Maybe it indicates a deeper trend or phenomenon. Many of the most influential scientific
discoveries were made when the investigator noticed an unexpected result, a value that was
not predicted by the classical theory. Albert Einstein, Louis Pasteur, and others built their
careers on exactly this circumstance.
3.3 Descriptive Statistics
3.3.1 Frequencies and Relative Frequencies
These are used for categorical data. The idea is that there are a number of different categories, and
we would like to get some idea about how the categories are represented in the population. For
example, we may want to see how the
3.3.2 Measures of Center
The sample mean is denoted x (read “x-bar”) and is simply the arithmetic average of the observa-
tions:
_ X\ + X2 + ■ ■ ■ + Xn
1
n
Z
Xi.
X =
n
(3.3.1)
36
CHAPTER 3. DATA DESCRIPTION
• Good: natural, easy to compute, has nice mathematical properties
• Bad: sensitive to extreme values
It is appropriate for use with data sets that are not highly skewed without extreme observations.
The sample median is another popular measure of center and is denoted x. To calculate its
value, first sort the data into an increasing sequence of numbers. If the data set has an odd number
of observations then x is the value of the middle observation, which lies in position (n + l)/2;
otherwise, there are two middle observations and x is the average of those middle values.
• Good: resistant to extreme values, easy to describe
• Bad: not as mathematically tractable, need to sort the data to calculate
One desirable property of the sample median is that it is resistant to extreme observations, in the
sense that the value of x depends only the values of the middle observations, and is quite unaf-
fected by the actual values of the outer observations in the ordered list. The same cannot be said
for the sample mean. Any significant changes in the magnitude of an observation Xk results in a
corresponding change in the value of the mean. Hence, the sample mean is said to be sensitive to
extreme observations.
The trimmed mean is a measure designed to address the sensitivity of the sample mean to
extreme observations. The idea is to “trim” a fraction (less than 1/2) of the observations off each
end of the ordered list, and then calculate the sample mean of what remains. We will denote it by
•*7=0.05-
• Good: resistant to extreme values, shares nice statistical properties
• Bad: need to sort the data
3.3.3 How to do it with R
• You can calculate frequencies or relative frequencies with the table function, and relative
frequencies with prop, table (table ()).
• You can calculate the sample mean of a data vector x with the command mean(x) .
• You can calculate the sample median of x with the command median(x) .
• You can calculate the trimmed mean with the trim argument; mean(x , trim = ® . 85) .
3.3.4 Order Statistics and the Sample Quantiles
A common first step in an analysis of a data set is to sort the values. Given a data set X\, x?, . . . ,xn,
we may sort the values to obtain an increasing sequence
*(D < X(2) < x(3) < • < x(n) (3.3.2)
and the resulting values are called the order statistics. The kth entry in the list, X(k), is the klh order
statistic, and approximately 100 (k/n)% of the observations fall below x^)- The order statistics give
3.3. DESCRIPTIVE STATISTICS
37
an indication of the shape of the data distribution, in the sense that a person can look at the order
statistics and have an idea about where the data are concentrated, and where they are sparse.
The sample quantiles are related to the order statistics. Unfortunately, there is not a universally
accepted definition of them. Indeed, R is equipped to calculate quantiles using nine distinct defini-
tions! We will describe the default method ( type = 7), but the interested reader can see the details
for the other methods with ?quantile.
Suppose the data set has n observations. Find the sample quantile of order p (0 < p < 1),
denoted qp , as follows:
First step: sort the data to obtain the order statistics x<\)y X(2), . ■ . ,X(„).
Second step: calculate (n - 1 )p + 1 and write it in the form k.d, where k is an integer and d is a
decimal.
Third step: The sample quantile qp is
kip = *(*) + d(x(k+ 1) - X(k)).
(3.3.3)
The interpretation of qp is that approximately 100 p% of the data fall below the value qp .
Keep in mind that there is not a unique definition of percentiles, quartiles, etc. Open a different
book, and you’ll find a different procedure. The difference is small and seldom plays a role except
in small data sets with repeated values. In fact, most people do not even notice in common use.
Clearly, the most popular sample quantile is §0.50, also known as the sample median, x. The
closest runners-up are the first quartile §0.25 and the third quartile §0.75 (the second quartile is the
median).
3.3.5 How to do it with R
At the command prompt We can find the order statistics of a data set stored in a vector x with
the command sort(x).
You can calculate the sample quantiles of any order p where 0 < p < 1 for a data set stored
in a data vector x with the quantile function, for instance, the command quantile (x , probs
= c(0, 8.25, 0. 37)) will return the smallest observation, the first quartile, §0.25. and the 37th
sample quantile, §0.37- F°r qp simply change the values in the probs argument to the value p.
With the R Commander In Rcmdr we can find the order statistics of a variable in the Active
data set by doing Data > Manage variables in Active data set. . . > Compute new variable
In the Expression to compute dialog simply type sort(varname), where varname is the variable
that it is desired to sort.
In Rcmdr, we can calculate the sample quantiles for a particular variable with the sequence
Statistics > Summaries > Numerical Summaries We can automatically calculate the quartiles
for all variables in the Active data set with the sequence Statistics > Summaries > Active
Dataset.
38
CHAPTER 3. DATA DESCRIPTION
3.3.6 Measures of Spread
Sample Variance and Standard Deviation The sample variance is denoted s2 and is calculated
with the formula
S2^Vfc-t)2. (3.3.4)
n — 1
i=i
The sample standard deviation is s = Vs2. Intuitively, the sample variance is approximately the
average squared distance of the observations from the sample mean. The sample standard deviation
is used to scale the estimate back to the measurement units of the original data.
• Good: tractable, has nice mathematical/statistical properties
• Bad: sensitive to extreme values
We will spend a lot of time with the variance and standard deviation in the coming chapters. In the
meantime, the following two rules give some meaning to the standard deviation, in that there are
bounds on how much of the data can fall past a certain distance from the mean.
Fact 3.12. Chebychev’s Rule: The proportion of observations within k standard deviations of the
mean is at least 1 - l/k2, i.e., at least 75%, 89%, and 94% of the data are within 2, 3, and 4
standard deviations of the mean, respectively.
Note that Chebychev’s Rule does not say anything about when k = 1, because 1 - l/l2 = 0,
which states that at least 0% of the observations are within one standard deviation of the mean
(which is not saying much).
Chebychev’s Rule applies to any data distribution, any list of numbers, no matter where it came
from or what the histogram looks like. The price for such generality is that the bounds are not very
tight; if we know more about how the data are shaped then we can say more about how much of
the data can fall a given distance from the mean.
Fact 3.13. Empirical Rule: If data follow a bell-shaped curve, then approximately 68%, 95%, and
99.7% of the data are within 1, 2, and 3 standard deviations of the mean, respectively.
Interquartile Range Just as the sample mean is sensitive to extreme values, so the associated
measure of spread is similarly sensitive to extremes. Further, the problem is exacerbated by the
fact that the extreme distances are squared. We know that the sample quartiles are resistant to
extremes, and a measure of spread associated with them is the interquartile range ( IQR ) defined
by IQR = <70.75 ~ <70.25 •
• Good: stable, resistant to outliers, robust to nonnormality, easy to explain
• Bad: not as tractable, need to sort the data, only involves the middle 50% of the data.
Median Absolute Deviation A measure even more robust than the IQR is the median absolute
deviation (MAD). To calculate it we first get the median ~x, next the absolute deviations \x\ - x\,
1*2 — x\, . . . , \x„ - x\, and the MAD is proportional to the median of those deviations:
MAD oc median(|*i - x\, |*2 - x\, . . . , \xn - *|).
(3.3.5)
3.3. DESCRIPTIVE STATISTICS
39
That is, the MAD = c ■ median(|xi - x\, \x2 - x\, . . . , \xn - Jc|), where c is a constant chosen so that
the MAD has nice properties. The value of c in R is by default c = 1.4286. This value is chosen to
ensure that the estimator of <x is correct, on the average, under suitable sampling assumptions (see
Section 9.1).
• Good: stable, very robust, even more so than the IQR.
• Bad: not tractable, not well known and less easy to explain.
Comparing Apples to Apples
We have seen three different measures of spread which, for a given data set, will give three different
answers. Which one should we use? It depends on the data set. If the data are well behaved, with
an approximate bell-shaped distribution, then the sample mean and sample standard deviation are
natural choices with nice mathematical properties. However, if the data have an unusual or skewed
shape with several extreme values, perhaps the more resistant choices among the IQR or MAD
would be more appropriate.
However, once we are looking at the three numbers it is important to understand that the esti-
mators are not all measuring the same quantity, on the average. In particular, it can be shown that
when the data follow an approximately bell-shaped distribution, then on the average, the sample
standard deviation s and the MAD will be the approximately the same value, namely, cr, but the
IQR will be on the average 1.349 times larger than s and the MAD. See 8 for more details.
3.3.7 How to do it with R
At the command prompt From the console we may compute the sample range with range (x)
and the sample variance with var (x) , where x is a numeric vector. The sample standard deviation
is sqrt(var(x)) or just sd(x). The IQR is IQR(x) and the median absolute deviation is mad(x).
In R Commander In Rcmdr we can calculate the sample standard deviation with the Statistics >
Summaries > Numerical Summaries. . . combination. R Commander does not calculate the IQR
or MAD in any of the menu selections, by default.
3.3.8 Measures of Shape
Sample Skewness The sample skewness , denoted by g i, is defined by the formula
_ 1 2"=1 (Xi-X)3
(3.3.6)
The sample skewness can be any value -oo < gi < oo. The sign of gi indicates the direction
of skewness of the distribution. Samples that have g\ > 0 indicate right-skewed distributions
(or positively skewed), and samples with g! < 0 indicate left-skewed distributions (or negatively
skewed). Values of gi near zero indicate a symmetric distribution. These are not hard and fast rules,
however. The value of gi is subject to sampling variability and thus only provides a suggestion to
the skewness of the underlying distribution.
We still need to know how big is “big”, that is, how do we judge whether an observed value
of gi is far enough away from zero for the data set to be considered skewed to the right or left? A
40
CHAPTER 3. DATA DESCRIPTION
good rule of thumb is that data sets with skewness larger than 2 sJ6/n in magnitude are substantially
skewed, in the direction of the sign of gi. See Tabachnick & Fidell [83] for details.
Sample Excess Kurtosis The sample excess kurtosis, denoted by gi, is given by the formula
82
l Z'Ute-^)4
3.
(3.3.7)
The sample excess kurtosis takes values -2 < g2 < °o. The subtraction of 3 may seem mysterious
but it is done so that mound shaped samples have values of g2 near zero. Samples with go > 0 are
called leptokurtic , and samples with go < 0 are called platykurtic. Samples with go ~ 0 are called
mesokurtic.
As a rule of thumb, if \g2\ >4 \j6/n then the sample excess kurtosis is substantially different
from zero in the direction of the sign of go- See Tabachnick & Fidell [83] for details.
Notice that both the sample skewness and the sample kurtosis are invariant with respect to
location and scale, that is, the values of gi and go do not depend on the measurement units of the
data.
3.3.9 How to do it with R
The el071 package [22] has the skewness function for the sample skewness and the kurtosis
function for the sample excess kurtosis. Both functions have a na . rm argument which is FALSE by
default.
Example 3.14. We said earlier that the discoveries data looked positively skewed; let’s see what
the statistics say:
> library (elQ71)
> skewness (discoveries)
[1] 1.20760®
> 2 * sqrt(6/length(discoveries))
[1] 0.4898979
The data are definitely skewed to the right. Let us check the sample excess kurtosis of the
UKDriverDeaths data:
> kurtosis(UKDriverDeaths)
[1] 0.07133848
> 4 * sqrt(6/length(UKDriverDeaths))
[1] 0.7071068
so that the UKDriverDeaths data appear to be mesokurtic, or at least not substantially lep-
tokurtic.
3.4. EXPLORATORY DATA ANALYSIS
41
3.4 Exploratory Data Analysis
This field was founded (mostly) by John Tukey (1915-2000). Its tools are useful when not much is
known regarding the underlying causes associated with the data set, and are often used for checking
assumptions. For example, suppose we perform an experiment and collect some data. . . now what?
We look at the data using exploratory visual tools.
3.4.1 More About Stemplots
There are many bells and whistles associated with stemplots, and the stem, leaf function can do
many of them.
Trim Outliers: Some data sets have observations that fall far from the bulk of the other data (in
a sense made more precise in Section 3.4.6). These extreme observations often obscure the
underlying structure to the data and are best left out of the data display. The trim, outliers
argument (which is TRUE by default) will separate the extreme observations from the others
and graph the stemplot without them; they are listed at the bottom (respectively, top) of the
stemplot with the label HI (respectively LO).
Split Stems: The standard stemplot has only one line per stem, which means that all observations
with first digit 3 are plotted on the same line, regardless of the value of the second digit. But
this gives some stemplots a “skyscraper” appearance, with too many observations stacked
onto the same stem. We can often fix the display by increasing the number of lines available
for a given stem. For example, we could make two lines per stem, say, 3* and 3 . . Obser-
vations with second digit 0 through 4 would go on the upper line, while observations with
second digit 5 through 9 would go on the lower line. (We could do a similar thing with five
lines per stem, or even ten lines per stem.) The end result is a more spread out stemplot
which often looks better. A good example of this was shown on page 34.
Depths: these are used to give insight into the balance of the observations as they accumulate
toward the median. In a column beside the standard stemplot, the frequency of the stem
containing the sample median is shown in parentheses. Next, frequencies are accumulated
from the outside inward, including the outliers. Distributions that are more symmetric will
have better balanced depths on either side of the sample median.
3.4.2 How to do it with R
At the command prompt The basic command is stem(x) or a more sophisticated version writ-
ten by Peter Wolf called stem, leaf (x) in the R Commander. We will describe stem, leaf since
that is the one used by R Commander.
With the R Commander WARNING: Sometimes when making a stem plot the result will not
be what you expected. There are several reasons for this:
• Stemplots by default will trim extreme observations (defined in Section 3.4.6) from the dis-
play. This in some cases will result in stemplots that are not as wide as expected.
42
CHAPTER 3. DATA DESCRIPTION
• The leafs digit is chosen automatically by stem. leaf according to an algorithm that the
computer believes will represent the data well. Depending on the choice of the digit, stem . leaf
may drop digits from the data or round the values in unexpected ways.
Let us take a look at the rivers data set.
> stem. leaf (rivers)
1 | 2: represents 120
leaf unit: 10
n: 141
1
1
1
3
29
2
1
0111133334555556666778888899
64
3
1
00000111122223333455555666677888999
(18)
4
1
011222233344566679
59
5
1
000222234467
47
6
1
0000112235789
34
7
1
12233368
26
8
1
04579
21
9
1
0008
17
10
1
035
14
11
1
07
12
12
1
047
9
13
1
0
HI: 1450 1459 1770 1885 2315 2348 2533 3710
The stemplot shows a right-skewed shape to the rivers data distribution. Notice that the last
digit of each of the data values were dropped from the display. Notice also that there were eight
extreme observations identified by the computer, and their exact values are listed at the bottom
of the stemplot. Look at the scale on the left of the stemplot and try to imagine how ridiculous
the graph would have looked had we tried to include enough stems to include these other eight
observations; the stemplot would have stretched over several pages. Notice finally that we can use
the depths to approximate the sample median for these data. The median lies in the row identified
by ( 18) , which means that the median is the average of the ninth and tenth observation on that row.
Those two values correspond to 43 and 43, so a good guess for the median would be 430. (For the
record, the sample median is ~x — 425. Recall that stemplots round the data to the nearest stem-leaf
pair.)
Next let us see what the precip data look like.
> stem. leaf (precip)
1 | 2: represents 12
leaf unit: 1
n: 70
L0: 7 7.2 7.8 7.8
8 1* | 1344
13 1. | 55677
16 2* | 024
18 2. | 59
3.4. EXPLORATORY DATA ANALYSIS
43
28
3*
| 0000111234
(15)
3.
I 555566677788899
27
4*
| 0000122222334
14
4.
I 56688899
6
5*
| 44
4
5.
I 699
HI: 67
Here is an example of split stems, with two lines per stem. The final digit of each datum has
been dropped for the display. The data appear to be left skewed with four extreme values to the
left and one extreme value to the right. The sample median is approximately 37 (it turns out to be
36.6).
3.4.3 Hinges and the Five Number Summary
Given a data set xi, X2, ■ ■ ■ , x„, the hinges are found by the following method:
• Find the order statistics X(ij, X(2), . . . , X(„).
• The lower hinge hi is in position L = \_(n + 3)/2J /2, where the symbol |_xj denotes the
largest integer less than or equal to x. If the position L is not an integer, then the hinge h; is
the average of the adjacent order statistics.
• The upper hinge hu is in position n + 1 — L.
Given the hinges, the five number summary ( 5NS ) is
5NS = (x(i), hL, x, hv, x(„)). (3.4.1)
An advantage of the 5 NS is that it reduces a potentially large data set to a shorter list of only five
numbers, and further, these numbers give insight regarding the shape of the data distribution similar
to the sample quantiles in Section 3.3.4.
3.4.4 How to do it with R
If the data are stored in a vector x, then you can compute the 5 NS with the fivenum function.
3.4.5 Boxplots
A boxplot is essentially a graphical representation of the 5 NS . It can be a handy alternative to a
stripchart when the sample size is large.
A boxplot is constructed by drawing a box alongside the data axis with sides located at the
upper and lower hinges. A line is drawn parallel to the sides to denote the sample median. Lastly,
whiskers are extended from the sides of the box to the maximum and minimum data values (more
precisely, to the most extreme values that are not potential outliers, defined below).
Boxplots are good for quick visual summaries of data sets, and the relative positions of the
values in the 5 NS are good at indicating the underlying shape of the data distribution, although
perhaps not as effectively as a histogram. Perhaps the greatest advantage of a boxplot is that it can
help to objectively identify extreme observations in the data set as described in the next section.
44
CHAPTER 3. DATA DESCRIPTION
Boxplots are also good because one can visually assess multiple features of the data set simul-
taneously:
Center can be estimated by the sample median, x.
Spread can be judged by the width of the box, hv - hL. We know that this will be close to the
IQR, which can be compared to s and the MAD , perhaps after rescaling if appropriate.
Shape is indicated by the relative lengths of the whiskers, and the position of the median inside the
box. Boxes with unbalanced whiskers indicate skewness in the direction of the long whisker.
Skewed distributions often have the median tending in the opposite direction of skewness.
Kurtosis can be assessed using the box and whiskers. A wide box with short whiskers will
tend to be platykurtic, while a skinny box with wide whiskers indicates leptokurtic distribu-
tions.
Extreme observations are identified with open circles (see below).
3.4.6 Outliers
A potential outlier is any observation that falls beyond 1 .5 times the width of the box on either side,
that is, any observation less than hi - 1.5 {hu - hi) or greater than hu + 1.5 (hu - hi). A suspected
outlier is any observation that falls beyond 3 times the width of the box on either side. In R, both
potential and suspected outliers (if present) are denoted by open circles; there is no distinction
between the two.
When potential outliers are present, the whiskers of the boxplot are then shortened to extend to
the most extreme observation that is not a potential outlier. If an outlier is displayed in a boxplot,
the index of the observation may be identified in a subsequent plot in Rcmdr by clicking the Identify
outliers with mouse option in the Boxplot dialog.
What do we do about outliers? They merit further investigation. The primary goal is to deter-
mine why the observation is outlying, if possible. If the observation is a typographical error, then
it should be corrected before continuing. If the observation is from a subject that does not belong
to the population of interest, then perhaps the datum should be removed. Otherwise, perhaps the
value is hinting at some hidden structure to the data.
3.4.7 How to do it with R
The quickest way to visually identify outliers is with a boxplot, described above. Another way is
with the boxplot . stats function.
Example 3.15. The rivers data. We will look for potential outliers in the rivers data.
> boxplot. stats (rivers) $out
[1] 1459 145® 1243 2348 371® 2315 2533 1306 127® 1885 177®
We may change the coef argument to 3 (it is 1.5 by default) to identify suspected outliers.
> boxplot . stats(rivers , coef = 3)$out
[1] 2348 371® 2315 2533 1885
3.5. MULTIVARIATE DATA AND DATA FRAMES
45
3.4.8 Standardizing variables
It is sometimes useful to compare data sets with each other on a scale that is independent of the
measurement units. Given a set of observed data X\ , xi, ■ ■ ■ , x„ we get z scores, denoted zi, Z/i, . . . ,
zn, by means of the following formula
Zi =
i — 1,2, . . . , n.
3.4.9 How to do it with R
The scale function will rescale a numeric vector (or data frame) by subtracting the sample mean
from each value (column) and/or by dividing each observation by the sample standard deviation.
3.5 Multivariate Data and Data Frames
We have had experience with vectors of data, which are long lists of numbers. Typically, each
entry in the vector is a single measurement on a subject or experimental unit in the study. We saw
in Section 2.3.3 how to form vectors with the c function or the scan function.
However, statistical studies often involve experiments where there are two (or more) measure-
ments associated with each subject. We display the measured information in a rectangular array
in which each row corresponds to a subject, and the columns contain the measurements for each
respective variable. For instance, if one were to measure the height and weight and hair color of
each of 1 1 persons in a research study, the information could be represented with a rectangular
array. There would be 1 1 rows. Each row would have the person’s height in the first column and
hair color in the second column.
The corresponding objects in R are called data frames, and they can be constructed with the
data, frame function. Each row is an observation, and each column is a variable.
Example 3.16. Suppose we have two vectors x and y and we want to make a data frame out of
them.
> x <- 5:8
> y <- letters[3:6]
> A <- data. frame(vl - x, v2 - y)
Notice that x and y are the same length. This is necessary. Also notice that x is a numeric
vector and y is a character vector. We may choose numeric and character vectors (or even factors)
for the columns of the data frame, but each column must be of exactly one type. That is, we can
have a column for height and a column for gender, but we will get an error if we try to mix
function height (numeric) and gender (character or factor) information in the same column.
Indexing of data frames is similar to indexing of vectors. To get the entry in row i and column
j do A [i , j ] . We can get entire rows and columns by omitting the other index.
> A[3, J
vl v2
3 7 e
46
CHAPTER 3. DATA DESCRIPTION
> A[l, J
vl v2
1 5 c
> A[, 2 ]
[1] c d e £
Levels: c d e £
There are several things happening above. Notice that A [ 3 , ] gave a data frame (with the same
entries as the third row of A) yet A [ 1 , ] is a numeric vector. A [ , 2 ] is a factor vector because the
default setting for data . frame is stringsAsFactors = TRUE.
Data frames have a names attribute and the names may be extracted with the names function.
Once we have the names we may extract given columns by way of the dollar sign.
> names (A)
[1] "vl" "v2"
> ASvl
[1] 5 6 7 8
The above is identical to A [ , 1] .
3.5.1 Bivariate Data
• Stacked bar charts
• odds ratio and relative risk
• Introduce the sample correlation coefficient.
The sample Pearson product-moment correlation coefficient:
r _ EILi (*; - *)0< - y)
r~ VE?=ife-?)VE?=i (yi-y)
• independent of scale
• -1 < r < 1
• measures strength and direction of linear association
• Two-Way Tables. Done with table, or in the R Commander by following Statistics > Con-
tingency Tables > Two-way Tables. You can also enter and analyze a two-way table.
o table
o prop. table
o addmargins
o rowPercents (Rcmdr)
3.6. COMPARING POPULATIONS
47
o colPercents (Rcmdr)
o totPercents(Rcmdr)
o A <- xtabs(~ gender + race, data = RcmdrTestDrive)
o xtabs( Freq ~ Class + Sex, data = Titanic) # from built in table
o barplot(A, legend. text=TRUE)
o barplot(t(A), legend. text=TRUE)
o barplot(A, legend. text=TRUE, beside = TRUE)
o spincplotigendcr ~ race, data = RcmdrTestDrive)
o Spine plot: plots categorical versus categorical
• Scatterplot: look for linear association and correlation.
o carb ~ optden, data = Formaldehyde (boring)
o cone ~ rate, data = Puromycin
o xyplot(accel ~ dist, data = attenu) nonlinear association
o xyplot(eruptions ~ waiting, data = faithful) (linear, two groups)
o xyplot(Petal. Width - Petal.Length, data = iris)
o xyplot(pressure ~ temperature, data = pressure) (exponential growth)
o xyplot(weight - height, data = women) (strong positive linear)
3.5.2 Multivariate Data
Multivariate Data Display
• Multi-Way Tables. You can do this with table, or in R Commander by following Statistics
> Contingency Tables > Multi-way Tables.
• Scatterplot matrix, used for displaying pairwise scatterplots simultaneously. Again, look for
linear association and correlation.
• 3D Scatterplot. See Figure 289
• plot(state. region, state . division)
• barplot(table(state. division, state. region) , legend. text=TRUE)
3.6 Comparing Populations
Sometimes we have data from two or more groups (or populations) and we would like to compare
them and draw conclusions. Some issues that we would like to address:
• Comparing centers and spreads: variation within versus between groups
• Comparing clusters and gaps
• Comparing outliers and unusual features
• Comparing shapes.
48
CHAPTER 3. DATA DESCRIPTION
3.6.1 Numerically
I am thinking here about the Statistics > Numerical Summaries > Summarize by groups option
or the Statistics > Summaries >Table of Statistics option.
3.6.2 Graphically
• Boxplots
o Variable width: the width of the drawn boxplots are proportional to y/nj, where n,- is
the size of the /th group. Why? Because many statistics have variability proportional to
the reciprocal of the square root of the sample size.
o Notches: extend to 1.58 • (hjj - hi)/ yfn. The idea is to give roughly a 95% confidence
interval for the difference in two medians. See Chapter 10.
• Stripcharts
o stripchart(weight ~ feed, method="stack", data=chickwts)
• Bar Graphs
o barplot(xtabs(Freq ~ Admit + Gender, data = UCBAdmissions)) # stacked bar chart
o barplot(xtabs(Freq ~ Admit, data = UCBAdmissions))
o barplot(xtabs(Freq ~ Gender + Admit, data = UCBAdmissions), legend = TRUE, be-
side = TRUE) # oops, discrimination.
o barplot(xtabs(Freq ~ Admit+Dept, data = UCBAdmissions), legend = TRUE, beside =
TRUE) # different departments have different standards
o barplot(xtabs(Freq ~ Gender+Dept, data = UCBAdmissions), legend = TRUE, beside
= TRUE) # men mostly applied to easy departments, women mostly applied to difficult
departments
o barplot(xtabs(Freq ~ Gender+Dept, data = UCBAdmissions), legend = TRUE, beside
= TRUE)
o barchartfAdmit ~ Freq, data = C)
o barchartfAdmit ~ Freq|Gender, data = C)
o barchartfAdmit ~ Freq | Dept, groups = Gender, data = C)
o barchartfAdmit ~ Freq | Dept, groups = Gender, data = C, auto. key = TRUE)
• Histograms
o ~ breaks | wool*tension, data = warpbreaks
o ~ weight | feed, data = chickwts
o ~ weight | group, data = PlantGrowth
o ~ count | spray, data = InsectSprays
o ~ len | dose, data = ToothGrowth
3.6. COMPARING POPULATIONS
49
o - decrease | treatment, data = OrchardSprays (or rowpos or colpos)
• Scatterplots
o xyplot(Petal. Width - Petal. Length, data = iris, group = Species)
> library(lattice)
> xyplot()
• Scatterplot matrices
o splom(~ cbind( GNP.deflator,GNP,Unemployed, Armed. Forces,Population, Year, Employed),
data = longley)
o splom(~ cbind(popl5,pop75,dpi), data = LifeCycleSavings)
o splom(~ cbind(Murder, Assault, Rape), data = USArrests)
o splom(~ cbind(CONT, INTG, DMNR), data = USJudgeRatings)
o splom(~ cbind(area,peri,shape,perm), data = rock)
o splom(~ cbind(Air.Flow, Water. Temp, Acid. Cone., stack.loss), data = stackloss)
o splom(~ cbind(Fertility,Agriculture,Examination,Education,Catholic,Infant.Mortality),
data = swiss)
o splom(~ cbind(Fertility,Agriculture,Examination), data = swiss) (positive and negative)
• Dot charts
o dotchart(U S PersonalExpenditure)
o dotchart(t)USPersonalExpenditure))
o dotchart(WorldPhones) (transpose is no good)
o freeny.x is no good, neither is volcano
o dotchart(UCBAdmissions[„l])
o dotplot(Survived ~ Freq | Class, groups = Sex, data = B)
o dotplot( Admit ~ Freq | Dept, groups = Gender, data = C)
• Mosaic plot
o mosaic)- Survived + Class + Age + Sex, data = Titanic) (or just mosaic(Titanic))
o mosaic)- Admit + Dept + Gender, data = UCBAdmissions)
• Spine plots
o spineplot(xtabs(Freq - Admit + Gender, data = UCBAdmissions)) # rescaled barplot
• Quantile-quantile plots: There are two ways to do this. One way is to compare two indepen-
dent samples (of the same size). qqplot(x,y). Another way is to compare the sample quantiles
of one variable to the theoretical quantiles of another distribution.
50
CHAPTER 3. DATA DESCRIPTION
Given two samples {xi, x2, . . . , xn} and {yi, y2, . . . , y„), we may find the order statistics X(\) <
X(2) < ■ ■ ■ < x(n) and V(i) < y(2) < < )’(„>■ Next, plot the n points (xa),y(i)), (xa),ya))
. . ,(x(H), y(n]).
It is clear that if X(k) = }'(k) for all k - 1,2, ... , n, then we will have a straight line. It is also
clear that in the real world, a straight line is NEVER observed, and instead we have a scatterplot
that hopefully had a general linear trend. What do the rules tell us?
• If the v-intercept of the line is greater (less) than zero, then the center of the Y data is greater
(less) than the center of the X data.
• If the slope of the line is greater (less) than one, then the spread of the Y data is greater (less)
than the spread of the X data..
3.6.3 Lattice Graphics
The following types of plots are useful when there is one variable of interest and there is a factor in
the data set by which the variable is categorized.
It is sometimes nice to set lattice .options (default . theme = "col .whitebg")
Side by side boxplots
> library (lattice)
> bwplot(~weight / feed, data =
chickwts)
3.6. COMPARING POPULATIONS
51
100 200 300 400
meatmeal
soybean
sunflower
i i
i i
•
•
° f
a oo
casein
horsebean
linseed
•
i
i
•
i
i
i
_ i
i
i
i
i i
i i
•
i i
100 200 300 400 100 200 300 400
weight
Figure 3.6.1: Boxplots of weight by feed type in the chickwts data
Histograms
> histogram(~age / education, data = infert)
52
CHAPTER 3. DATA DESCRIPTION
CO
H — 1
O
O
-i — «
c=
<d
o
* —
CD
CL
20 25 30 35 40 45
age
Figure 3.6.2: Histograms of age by education level from the infert data
Scatterplots
> xyplot (Petal. Length ~ Petal .Width / Species
data
iris)
3.6. COMPARING POPULATIONS
53
virginica
- 7
- 6
- 5
- 4
- 3
- 2
- 1
setosa
versicolor
sP°
r
0.0 0.5 1.0 1.5 2.0 2.5
Petal. Width
Figure 3.6.3: An xyplot of Petal .Length versus Petal .Width by Species in the iris data
Coplots
> coplot (cone ~ uptake / Type * Treatment , data = C02)
54
CHAPTER 3. DATA DESCRIPTION
NULL
Given : Type
Mississippi
Quebec
o
c=
o
o
10 20 30 40
a
CD
E
cd
<d
c
a)
>
CD
uptake
Figure 3.6.4: A coplot of cone versus uptake by Type and Treatment in the C02 data
3.6. COMPARING POPULATIONS
55
Chapter Exercises
Directions: Open R and issue the following commands at the command line to get started. Note
that you need to have the RcmdrPlugin . IPSUR package installed, and for some exercises you need
the el071 package.
library (RcmdrPlugin. IPSUR)
data(RcmdrTestDrive)
attach(RcmdrTestDrive)
names (RcmdrTestDrive) # shows names of variables
To load the data in the R Commander (Rcmdr), click the Data Set button, and select RcmdrTestDrive
as the active data set. To learn more about the data set and where it comes from, type ?RcmdrTestDrive
at the command line.
Exercise 3.1. Perform a summary of all variables in RcmdrTestDrive. You can do this with the
command
summary (RcmdrTestDrive)
Alternatively, you can do this in the Rcmdr with the sequence Statistics > Summaries > Active
Data Set. Report the values of the summary statistics for each variable.
Answer:
> summary(RcmdrTestDrive)
order
race
smoke
gender
salary
Min.
1.00
AfAmer
18
No : 134
Female
95
Min.
11.62
1st Qu.
42.75
Asian
8
Yes: 34
Male
73
1st
Qu.
15.93
Median
84.50
Other
16
Median
17.59
Mean
84.50
White
126
Mean
17.10
3rd Qu.
126.25
3rd Qu.
18.46
Max.
168.00
Max
21.19
reduction
before
after
parking
Min.
4.904
Min.
51.
17
Min.
48.79
Min
1.000
1st Qu.
5.195
1st Qu.
63.
36
1st Qu.
62.80
1st
Qu.
1.000
Median
5.501
Median
67.
62
Median
66.94
Median
2.000
Mean
5.609
Mean
67.
36
Mean
66.85
Mean
2.524
3rd Qu.
5.989
3rd Qu.
71.
28
3rd Qu.
70.88
3rd
Qu.
3.000
Max.
6.830
Max.
89.
96
Max.
89.89
Max
18.000
Exercise 3.2. Make a table of the race variable. Do this with Statistics > Summaries > IPSUR -
Frequency Distributions...
1 . Which ethnicity has the highest frequency?
2. Which ethnicity has the lowest frequency?
3. Include a bar graph of race. Do this with Graphs > IPSUR - Bar Graph...
Frequency
56
CHAPTER 3. DATA DESCRIPTION
Solution: First we will make a table of the race variable with the table function.
> table (race)
race
AfAraer Asian Other White
18 8 16 126
1 . For these data. White has the highest frequency.
2. For these data, Asian has the lowest frequency.
3. The graph is shown below.
O
C\J
O
00
O
CD
o
o
CM
O -
AfAmer Asian Other White
race
Exercise 3.3. Calculate the average salary by the factor gender. Do this with Statistics > Sum-
maries > Table of Statistics...
1 . Which gender has the highest mean salary ?
2. Report the highest mean salary.
3. Compare the spreads for the genders by calculating the standard deviation of salary by gen-
der. Which gender has the biggest standard deviation?
3.6. COMPARING POPULATIONS
57
4. Make boxplots of salary by gender with the following method:
On the Rcmdr, click Graphs > IPSUR - Boxplot...
In the Variable box, select salary.
Click the Plot by groups... box and select gender. Click OK.
Click OK to graph the boxplot.
How does the boxplot compare to your answers to (1) and (3)7
Solution: We can generate a table listing the average salaries by gender with two methods. The
first uses tapply:
> x <- tapply (salary , list (gender = gender), mean)
> x
gender
Female Male
16.46353 17.93035
The second method uses the by function:
> by (salary , gender, mean, na.rm = TRUE)
gender: Female
[1] 16.46353
gender: Male
[1] 17.93035
Now to answer the questions:
1 . Which gender has the highest mean salary?
We can answer this by looking above. For these data, the gender with the highest mean salary
is Male.
2. Report the highest mean salary.
Depending on our answer above, we would do something like
meanfsalary [gender == Male])
for example. For these data, the highest mean salary is
> x[which(x == max(x))]
Male
17.93035
58
CHAPTER 3. DATA DESCRIPTION
3. Compare the spreads for the genders by calculating the standard deviation of salary by gen-
der. Which gender has the biggest standard deviation?
> y <- tapply (salary , list(gender = gender), sd)
> y
gender
Female Male
2.122113 1.077183
For these data, the the largest standard deviation is approximately 2.12 which was attained
by the Female gender.
4. Make boxplots of salary by gender. How does the boxplot compare to your answers to (1)
and (3)?
The graph is shown below.
0
"O
c:
0
O)
0
CCS
0
0
E
0
12
14
16
18
20
salary
3.6. COMPARING POPULATIONS
59
Answers will vary. There should be some remarks that the center of the box is farther to the
right for the Male gender, and some recognition that the box is wider for the Female gender.
Exercise 3.4. For this problem we will study the variable reduction.
1. Find the order statistics and store them in a vector x. Hint: x <- sort (reduction)
2. Find X(i37), the 137th order statistic.
3. Find the IQR.
4. Find the Five Number Summary (5NS).
5. Use the 5NS to calculate what the width of a boxplot of reduction would be.
6. Compare your answers (3) and (5). Are they the same? If not, are they close?
7. Make a boxplot of reduction, and include the boxplot in your report. You can do this with
the boxplot function, or in Rcmdr with Graphs > IPSUR - Boxplot...
8. Are there any potential/suspected outliers? If so, list their values. Hint: use your answer to
(a).
9. Using the rules discussed in the text, classify answers to (8), if any, as potential or suspected
outliers.
Answers:
> x[137]
[1] 6.101618
> IQR (x)
[1] 0.7943932
> fivenum(x)
[1] 4.903922 5.193638 5.501241 5.989846 6.830096
> fi venum (x) [4] - fi venum(x) [2]
[1] 0.796208
Compare your answers (3) and (5). Are they the same? If not, are they close?
Yes, they are close, within 0.00181484542950905 of each other.
The boxplot of reduction is below.
60
CHAPTER 3. DATA DESCRIPTION
5.0 5.5 6.0 6.5
reduction
> temp <- fivenum(x)
> inF <- 1.5 * ( temp[4] - temp[2]) + temp[4]
> outF <- 3 * (temp[4] - temp[2J) + temp[4]
> which (x > inF)
integer(Q)
> which (x > outF)
integer(Q)
Observations would be considered potential outliers, while observation(s) would be considered
a suspected outlier.
Exercise 3.5. In this problem we will compare the variables before and after. Don’t forget library (e 107 1) .
1. Examine the two measures of center for both variables. Judging from these measures, which
variable has a higher center?
2. Which measure of center is more appropriate for before ? (You may want to look at a boxplot.)
Which measure of center is more appropriate for after ?
3. Based on your answer to (2), choose an appropriate measure of spread for each variable,
calculate it, and report its value. Which variable has the biggest spread? (Note that you need
to make sure that your measures are on the same scale.)
3.6. COMPARING POPULATIONS
61
4. Calculate and report the skewness and kurtosis for before. Based on these values, how would
you describe the shape of before !
5. Calculate and report the skewness and kurtosis for after. Based on these values, how would
you describe the shape of after ?
6. Plot histograms of before and after and compare them to your answers to (4) and (5).
Solution:
1 . Examine the two measures of center for both variables that you found in problem 1 . Judging
from these measures, which variable has a higher center?
We may take a look at the summary (RcmdrTestDrive) output from Exercise 3.1. Here we
will repeat the relevant summary statistics.
> c(mean(before) , median (be fore))
[1] 67.36338 67.61824
> c(mean(after) , median (after))
[1] 66.85215 66.93688
The idea is to look at the two measures and compare them to make a decision. In a nice
world, both the mean and median of one variable will be larger than the other which sends a
nice message. If We get a mixed message, then we should look for other information, such
as extreme values in one of the variables, which is one of the reasons for the next part of the
problem.
2. Which measure of center is more appropriate for before ? (You may want to look at a boxplot.)
Which measure of center is more appropriate for after ?
The boxplot of before is shown below.
62
CHAPTER 3. DATA DESCRIPTION
before
We want to watch out for extreme values (shown as circles separated from the box) or large
departures from symmetry. If the distribution is fairly symmetric then the mean and median
should be approximately the same. But if the distribution is highly skewed with extreme
values then we should be skeptical of the sample mean, and fall back to the median which
is resistant to extremes. By design, the before variable is set up to have a fairly symmetric
distribution.
A boxplot of after is shown next.
3.6. COMPARING POPULATIONS
63
1
50
1
60
1
70
1
80
1
90
after
The same remarks apply to the after variable. The after variable has been designed to be
left-skewed. . . thus, the median would likely be a good choice for this variable.
3. Based on your answer to (2), choose an appropriate measure of spread for each variable,
calculate it, and report its value. Which variable has the biggest spread? (Note that you need
to make sure that your measures are on the same scale.)
Since before has a symmetric, mound shaped distribution, an excellent measure of center
would be the sample standard deviation. And since after is left-skewed, we should use the
median absolute deviation. It is also acceptable to use the IQR, but we should rescale it
appropriately, namely, by dividing by 1.349. The exact values are shown below.
> sd (be fore )
[1] 6 . 201724
> mad(after)
[1] 6.095189
64
CHAPTER 3. DATA DESCRIPTION
> IQR(after)/l . 349
[1] 5.986954
Judging from the values above, we would decide which variable has the higher spread. Look
at how close the mad and the IQR (after suitable rescaling) are; it goes to show why the
rescaling is important.
4. Calculate and report the skewness and kurtosis for before. Based on these values, how would
you describe the shape of before ?
The values of these descriptive measures are shown below.
> library (el671)
> skewness (before)
[1] 8.4016912
> kurtosis (be fore)
[1] 1.542225
We should take the sample skewness value and compare it to 2 y/6/n *0.378 in absolute
value to see if it is substantially different from zero. The direction of skewness is decided by
the sign (positive or negative) of the skewness value.
We should take the sample kurtosis value and compare it to 2 • V24/168 *0.756), in absolute
value to see if the excess kurtosis is substantially different from zero. And take a look at the
sign to see whether the distribution is platykurtic or leptokurtic.
5. Calculate and report the skewness and kurtosis for after. Based on these values, how would
you describe the shape of after ?
The values of these descriptive measures are shown below.
> skewness (after)
[1] 8.3235134
> kurtosis (after)
[1] 1.452381
We should do for this one just like we did previously. We would again compare the sample
skewness and kurtosis values (in absolute value) to 0.378 and 0.756, respectively.
6. Plot histograms of before and after and compare them to your answers to (4) and (5).
The graphs are shown below.
Frequency
3.6. COMPARING POPULATIONS
65
Histogram of before
50 60 70 80 90
before
66
CHAPTER 3. DATA DESCRIPTION
Histogram of after
>
o
a
CD
=3
c r
CD
o
50 60 70
80
90
after
Answers will vary. We are looking for visual consistency in the histograms to our statements
above.
Exercise 3.6. Describe the following data sets just as if you were communicating with an alien,
but one who has had a statistics class. Mention the salient features (data type, important proper-
ties, anything special). Support your answers with the appropriate visual displays and descriptive
statistics.
1. Conversion rates of Euro currencies stored in euro.
2. State abbreviations stored in state . abb.
Chapter 4
Probability
In this chapter we define the basic terminology associated with probability and derive some of its
properties. We discuss three interpretations of probability. We discuss conditional probability and
independent events, along with Bayes’ Theorem. We finish the chapter with an introduction to
random variables, which paves the way for the next two chapters.
In this book we distinguish between two types of experiments: deterministic and random. A
deterministic experiment is one whose outcome may be predicted with certainty beforehand, such
as combining Hydrogen and Oxygen, or adding two numbers such as 2 + 3. A random experiment
is one whose outcome is determined by chance. We posit that the outcome of a random experiment
may not be predicted with certainty beforehand, even in principle. Examples of random experi-
ments include tossing a coin, rolling a die, and throwing a dart on a board, how many red lights
you encounter on the drive home, how many ants traverse a certain patch of sidewalk over a short
period, etc.
What do I want them to know?
• that there are multiple interpretations of probability, and the methods used depend somewhat
on the philosophy chosen
• nuts and bolts of basic probability jargon: sample spaces, events, probability functions, etc.
• how to count
• conditional probability and its relationship with independence
• Bayes’ Rule and how it relates to the subjective view of probability
• what we mean by ’random variables’, and where they come from
4.1 Sample Spaces
For a random experiment E, the set of all possible outcomes of E is called the sample space and
is denoted by the letter S . For a coin-toss experiment, S would be the results “Head” and “Tail”,
67
68
CHAPTER 4. PROBABILITY
Figure 4.0. 1 : Two types of experiments
which we may represent by S — {H, T). Formally, the performance of a random experiment is the
unpredictable selection of an outcome in S .
4.1.1 How to do it with R
Most of the probability work in this book is done with the prob package [52]. A sample space
is (usually) represented by a data frame , that is, a rectangular collection of variables (see Section
3.5.2). Each row of the data frame corresponds to an outcome of the experiment. The data frame
choice is convenient both for its simplicity and its compatibility with the R Commander. Data
frames alone are, however, not sufficient to describe some of the more interesting probabilistic
applications we will study later; to handle those we will need to consider a more general list data
structure. See Section 4.6.3 for details.
Example 4.1. Consider the random experiment of dropping a Styrofoam cup onto the floor from
a height of four feet. The cup hits the ground and eventually comes to rest. It could land upside
down, right side up, or it could land on its side. We represent these possible outcomes of the random
experiment by the following.
> S <- data, frame (lands = c("down", "up", "side"))
> S
lands
1 down
2 up
3 side
The sample space S contains the column lands which stores the outcomes "down", "up", and
"side".
4.1. SAMPLE SPACES
69
Some sample spaces are so common that convenience wrappers were written to set them up with
minimal effort. The underlying machinery that does the work includes the expand, grid function
in the base package, combn in the combinat package [14], and permsn in the prob package1.
Consider the random experiment of tossing a coin. The outcomes are H and T. We can set up
the sample space quickly with the tosscoin function:
> library (prob)
> tosscoin(l)
tossl
1 H
2 T
The number 1 tells tosscoin that we only want to toss the coin once. We could toss it three
times:
> tosscoin(3)
tossl toss2 toss3
1 H H H
2 T H H
3 H T H
4 T T H
5 H H T
6 T H T
7 H T T
8 T t i
Alternatively we could roll a fair die:
> rolldie(l)
XI
1 1
2 2
3 3
4 4
5 5
6 6
The rolldie function defaults to a 6-sided die, but we can specify others with the nsides
argument. The command rolldie (3, nsides = 4) would be used to roll a 4-sided die three
times.
Perhaps we would like to draw one card from a standard set of playing cards (it is a long data
frame):
> head(cardsO)
1 The seasoned R user can get the job done without the convenience wrappers. I encourage the beginner to use them
to get started, but I also recommend that introductory students wean themselves as soon as possible. The wrappers were
designed for ease and intuitive use, not for speed or efficiency.
70
CHAPTER 4. PROBABILITY
rank suit
1 2 Club
2 3 Club
3 4 Club
4 5 Club
5 6 Club
6 7 Club
The cards function that we just used has optional arguments jokers (if you would like Jokers
to be in the deck) and makespace which we will discuss later. There is also a roulette function
which returns the sample space associated with one spin on a roulette wheel. There are EU and
USA versions available. Interested readers may contribute any other game or sample spaces that
may be of general interest.
4.1.2 Sampling from Urns
This is perhaps the most fundamental type of random experiment. We have an urn that contains a
bunch of distinguishable objects (balls) inside. We shake up the urn, reach inside, grab a ball, and
take a look. That’s all.
But there are all sorts of variations on this theme. Maybe we would like to grab more than one
ball - say, two balls. What are all of the possible outcomes of the experiment now? It depends on
how we sample. We could select a ball, take a look, put it back, and sample again. Another way
would be to select a ball, take a look - but do not put it back - and sample again (equivalently, just
reach in and grab two balls). There are certainly more possible outcomes of the experiment in the
former case than in the latter. In the first (second) case we say that sampling is done with ( without )
replacement.
There is more. Suppose we do not actually keep track of which ball came first. All we observe
are the two balls, and we have no idea about the order in which they were selected. We call this
unordered sampling (in contrast to ordered ) because the order of the selections does not matter
with respect to what we observe. We might as well have selected the balls and put them in a bag
before looking.
Note that this one general class of random experiments contains as a special case all of the
common elementary random experiments. Tossing a coin twice is equivalent to selecting two balls
labeled H and T from an urn, with replacement. The die-roll experiment is equivalent to selecting
a ball from an urn with six elements, labeled 1 through 6.
4.1.3 How to do it with R
The prob package accomplishes sampling from urns with the urnsamples function, which has
arguments x, size, replace, and ordered. The argument x represents the urn from which sam-
pling is to be done. The size argument tells how large the sample will be. The ordered and
replace arguments are logical and specify how sampling will be performed. We will discuss each
in turn.
Example 4.2. Let our urn simply contain three balls, labeled 1, 2, and 3, respectively. We are
going to take a sample of size 2 from the urn.
4.1. SAMPLE SPACES
71
Ordered, With Replacement
If sampling is with replacement, then we can get any outcome 1, 2, or 3 on any draw. Further, by
“ordered” we mean that we shall keep track of the order of the draws that we observe. We can
accomplish this in R with
> urnsamples(l : 3 , size = 2, replace = TRUE, ordered = TRUE)
XI X2
1 1 1
2 2 1
3 3 1
4 1 2
5 2 2
6 3 2
7 1 3
8 2 3
9 3 3
Notice that rows 2 and 4 are identical, save for the order in which the numbers are shown.
Further, note that every possible pair of the numbers 1 through 3 are listed. This experiment is
equivalent to rolling a 3-sided die twice, which we could have accomplished with rolldie(2 ,
nsides = 3).
Ordered, Without Replacement
Here sampling is without replacement, so we may not observe the same number twice in any row.
Order is still important, however, so we expect to see the outcomes 1 , 2 and 2 , 1 somewhere in our
data frame.
> urnsamples(l : 3 , size = 2, replace = FALSE, ordered = TRUE)
XI X2
1 1 2
2 2 1
3 1 3
4 3 1
5 2 3
6 3 2
This is just as we expected. Notice that there are less rows in this answer due to the more
restrictive sampling procedure. If the numbers 1, 2, and 3 represented “Fred”, “Mary”, and “Sue”,
respectively, then this experiment would be equivalent to selecting two people of the three to serve
as president and vice-president of a company, respectively, and the sample space shown above lists
all possible ways that this could be done.
Unordered, Without Replacement
Again, we may not observe the same outcome twice, but in this case, we will only retain those
outcomes which (when jumbled) would not duplicate earlier ones.
72
CHAPTER 4. PROBABILITY
> urnsamples (1 : 3 , size = 2, replace = FALSE, ordered = FALSE)
XI X2
1 1 2
2 1 3
3 2 3
This experiment is equivalent to reaching in the urn, picking a pair, and looking to see what
they are. This is the default setting of urnsaraples, so we would have received the same output by
simply typing urnsamples (1 : 3 , 2).
Unordered, With Replacement
The last possibility is perhaps the most interesting. We replace the balls after every draw, but we
do not remember the order in which the draws came.
> urnsamples(l : 3 , size = 2, replace = TRUE, ordered = FALSE)
XI X2
1 1 1
2 1 2
3 1 3
4 2 2
5 2 3
6 3 3
We may interpret this experiment in a number of alternative ways. One way is to consider this
as simply putting two 3-sided dice in a cup, shaking the cup, and looking inside - as in a game
of Liar’s Dice, for instance. Each row of the sample space is a potential pair we could observe.
Another way is to view each outcome as a separate method to distribute two identical golf balls into
three boxes labeled 1, 2, and 3. Regardless of the interpretation, urnsamples lists every possible
way that the experiment can conclude.
Note that the urn does not need to contain numbers; we could have just as easily taken our
urn to be x = c("Red" , "Blue" , "Green"). But, there is an important point to mention before
proceeding. Astute readers will notice that in our example, the balls in the urn were distinguishable
in the sense that each had a unique label to distinguish it from the others in the urn. A natural
question would be, “What happens if your urn has indistinguishable elements, for example, what
if x = c("Red" , "Red" , "Blue")?” The answer is that urnsamples behaves as if each ball in
the urn is distinguishable, regardless of its actual contents. We may thus imagine that while there
are two red balls in the urn, the balls are such that we can tell them apart (in principle) by looking
closely enough at the imperfections on their surface.
In this way, when the x argument of urnsamples has repeated elements, the resulting sam-
ple space may appear to be ordered = TRUE even when, in fact, the call to the function was
urnsamples ( . . . , ordered = FALSE). Similar remarks apply for the replace argument.
4.2. EVENTS
73
4.2 Events
An event A is merely a collection of outcomes, or in other words, a subset of the sample space2.
After the performance of a random experiment E we say that the event A occurred if the experi-
ment’s outcome belongs to A. We say that a bunch of events A i, A 2, A3, . . . are mutually exclusive
or disjoint if A, Pi Aj = 0 for any distinct pair A, + A j. For instance, in the coin-toss experiment the
events A = {Heads) and B = {Tails) would be mutually exclusive. Now would be a good time to
review the algebra of sets in Appendix E. 1.
4.2.1 How to do it with R
Given a data frame sample/probability space S, we may extract rows using the [] operator:
> S <- tosscoin(2 , makespace = TRUE)
tossl toss2 probs
1 H H 8.25
2 T H 8.25
3 H T 8.25
4 T T 8.25
> S[l:3, J
tossl toss2 probs
1 H H 8.25
2 T H 8.25
3 H T 8.25
> S[c(2, 4), ]
tossl toss2 probs
2 T H 8.25
4 T T 8.25
and so forth. We may also extract rows that satisfy a logical expression using the subset
function, for instance
> S <- cards Q
> subset (S, suit == "Heart")
rank suit
27 2 Heart
28 3 Heart
29 4 Heart
38 5 Heart
31 6 Heart
-This naive definition works for finite or countably infinite sample spaces, but is inadequate for sample spaces in general.
In this book, we will not address the subtleties that arise, but will refer the interested reader to any text on advanced
probability or measure theory.
74
CHAPTER 4. PROBABILITY
32 7 Heart
33 8 Heart
34 9 Heart
35 1® Heart
36 J Heart
37 Q Heart
38 K Heart
39 A Heart
> subset (S, rank %in% 7:9)
rank suit
6 7 Club
7 8 Club
8 9 Club
19 7 Diamond
2® 8 Diamond
21 9 Diamond
32 7 Heart
33 8 Heart
34 9 Heart
45 7 Spade
46 8 Spade
47 9 Spade
We could continue indefinitely. Also note that mathematical expressions are allowed:
> subset (rolldie (3) , XI + X2 + X3 > 16)
XI X2 X3
18® 6 6 5
21® 6 5 6
215 5 6 6
216 6 6 6
4.2.2 Functions for Finding Subsets
It does not take long before the subsets of interest become complicated to specify. Yet the main
idea remains: we have a particular logical condition to apply to each row. If the row satisfies the
condition, then it should be in the subset. It should not be in the subset otherwise. The ease with
which the condition may be coded depends of course on the question being asked. Here are a few
functions to get started.
The %in% function
The function %in% helps to learn whether each value of one vector lies somewhere inside another
vector.
4.2. EVENTS
75
> x <- 1:16)
> y <- 8:12
> y %in% x
[1] TRUE TRUE TRUE FALSE FALSE
Notice that the returned value is a vector of length 5 which tests whether each element of y is
in x, in turn.
The isin function
It is more common to want to know whether the whole vector y is in x. We can do this with the
isin function.
> isin(x, y)
[1] FALSE
Of course, one may ask why we did not try something like all (y %in% x), which would give
a single result, TRUE. The reason is that the answers are different in the case that y has repeated
values. Compare:
> x <- 1:10
> y <- c(3, 3, 7)
> all(y %in% x)
[1] TRUE
> isin(x, y)
[1] FALSE
The reason for the above is of course that x contains the value 3, but x does not have two
3’s. The difference is important when rolling multiple dice, playing cards, etc. Note that there is
an optional argument ordered which tests whether the elements of y appear in x in the order in
which they are appear in y. The consequences are
> isin(x, c(3, 4, 5), ordered = TRUE)
[1] TRUE
> isin(x, c(3, 5, 4), ordered = TRUE)
[1] FALSE
The connection to probability is that have a data frame sample space and we would like to find a
subset of that space. A data . frame method was written for isin that simply applies the function
to each row of the data frame. We can see the method in action with the following:
> S <- rolldie(4)
> subset(S, isin(S, c(2, 2, 6), ordered = TRUE))
76
CHAPTER 4. PROBABILITY
188
XI
2
X2
2
X3
6
X4
1
404
2
2
6
2
620
2
2
6
3
836
2
2
6
4
1052
2
2
6
5
1088
2
2
1
6
1118
2
1
2
6
1123
1
2
2
6
1124
2
2
2
6
1125
3
2
2
6
1126
4
2
2
6
1127
5
2
2
6
1128
6
2
2
6
1130
2
3
2
6
1136
2
4
2
6
1142
2
5
2
6
1148
2
6
2
6
1160
2
2
3
6
1196
2
2
4
6
1232
2
2
5
6
1268
2
2
6
6
There are a few other functions written to find useful subsets, namely, countrep and isrep.
Essentially these were written to test for (or count) a specific number of designated values in out-
comes. See the documentation for details.
4.2.3 Set Union, Intersection, and Difference
Given subsets A and B, it is often useful to manipulate them in an algebraic fashion. To this end,
we have three set operations at our disposal: union, intersection, and difference. Below is a table
that summarizes the pertinent information about these operations.
Name
Denoted
Defined by elements
Code
Union
Intersection
Difference
AUB
AnB
A\B
in A or B or both
in both A and B
in A but not in B
union(A,B)
intersect(A,B)
setdiff(A,B)
Some examples follow.
> S = cards O
> A = subset (S, suit == "Heart")
> B = subsetCS, rank %in% 7:9)
We can now do some set algebra:
> union(A, B)
4.2. EVENTS
77
rank
suit
6
7
Club
7
8
Club
8
9
Club
19
7
Diamond
2®
8
Diamond
21
9
Diamond
27
2
Heart
28
3
Heart
29
4
Heart
B®
5
Heart
31
6
Heart
32
7
Heart
33
8
Heart
34
9
Heart
35
10
Heart
36
J
Heart
37
Q
Heart
38
K
Heart
39
A
Heart
45
7
Spade
46
8
Spade
47
9
Spade
> .
intersect (A, ,
rank
suit
32
7
Heart
33
8
Heart
34
9
Heart
> .
s etdiff(A, B)
rank
suit
27
2
Heart
28
3
Heart
29
4
Heart
3®
5
Heart
31
6
Heart
35
1®
Heart
36
J
Heart
37
Q
Heart
38
K
Heart
39
A
Heart
> .
s etdiff(B, A)
rank
suit
6
7
Club
7
8
Club
78
CHAPTER 4. PROBABILITY
8
9
Club
19
7
Diamond
20
8
Diamond
21
9
Diamond
45
7
Spade
46
8
Spade
47
9
Spade
Notice that setdiff is not symmetric. Further, note that we can calculate the complement of a
set A, denoted Ac and defined to be the elements of S that are not in A simply with setdiff (S , A) .
There have been methods written for intersect, setdiff, subset, and union in the case
that the input objects are of class ps. See Section 4.6.3.
Note 4.3. When the prob package loads you will notice a message: “The following object (s)
are masked from package: base : intersect, setdiff, union”. The reason for this mes-
sage is that there already exist methods for the functions intersect, setdiff, subset, and
union in the base package which ships with R. However, these methods were designed for when
the arguments are vectors of the same mode. Since we are manipulating sample spaces which are
data frames and lists, it was necessary to write methods to handle those cases as well. When the
prob package is loaded, R recognizes that there are multiple versions of the same function in the
search path and acts to shield the new definitions from the existing ones. But there is no cause for
alarm, thankfully, because the prob functions have been carefully defined to match the usual base
package definition in the case that the arguments are vectors.
4.3 Model Assignment
Let us take a look at the coin-toss experiment more closely. What do we mean when we say “the
probability of Heads” or write P(Heads)? Given a coin and an itchy thumb, how do we go about
finding what P(Heads) should be?
4.3.1 The Measure Theory Approach
This approach states that the way to handle P(Heads) is to define a mathematical function, called
a probability measure , on the sample space. Probability measures satisfy certain axioms (to be
introduced later) and have special mathematical properties, so not just any mathematical function
will do. But in any given physical circumstance there are typically all sorts of probability measures
from which to choose, and it is left to the experimenter to make a reasonable choice - usually based
on considerations of objectivity. For the tossing coin example, a valid probability measure assigns
probability p to the event {Heads), where p is some number 0 < p < 1. An experimenter that
wishes to incorporate the symmetry of the coin would choose p = 1/2 to balance the likelihood of
{Heads) and {Tails).
Once the probability measure is chosen (or determined), there is not much left to do. All
assignments of probability are made by the probability function, and the experimenter needs only to
plug the event (Heads) into to the probability function to find P(Heads). In this way, the probability
of an event is simply a calculated value, nothing more, nothing less. Of course this is not the whole
story; there are many theorems and consequences associated with this approach that will keep
4.3. MODEL ASSIGNMENT
19
us occupied for the remainder of this book. The approach is called measure theory because the
measure (probability) of a set (event) is associated with how big it is (how likely it is to occur).
The measure theory approach is well suited for situations where there is symmetry to the exper-
iment, such as flipping a balanced coin or spinning an arrow around a circle with well-defined pie
slices. It is also handy because of its mathematical simplicity, elegance, and flexibility. There are
literally volumes of information that one can prove about probability measures, and the cold rules
of mathematics allow us to analyze intricate probabilistic problems with vigor.
The large degree of flexibility is also a disadvantage, however. When symmetry fails it is
not always obvious what an “objective” choice of probability measure should be; for instance,
what probability should we assign to {Heads) if we spin the coin rather than flip it? (It is not
1/2.) Furthermore, the mathematical rules are restrictive when we wish to incorporate subjective
knowledge into the model, knowledge which changes over time and depends on the experimenter,
such as personal knowledge about the properties of the specific coin being flipped, or of the person
doing the flipping.
The mathematician who revolutionized this way to do probability theory was Andrey Kol-
mogorov, who published a landmark monograph in 1933. See
http://www-history.mcs . st-andrews . ac.uk/Mathematicians/Kolmogorov.html
for more information.
4.3.2 Relative Frequency Approach
This approach states that the way to determine P(Heads) is to flip the coin repeatedly, in exactly
the same way each time. Keep a tally of the number of flips and the number of Heads observed.
Then a good approximation to P(Heads) will be
m number of observed Heads
P(Heads) ~
total number of flips
(4.3.1)
The mathematical underpinning of this approach is the celebrated Law of Large Numbers,
which may be loosely described as follows. Let £ be a random experiment in which the event A
either does or does not occur. Perform the experiment repeatedly, in an identical manner, in such
a way that the successive experiments do not influence each other. After each experiment, keep
a running tally of whether or not the event A occurred. Let S„ count the number of times that A
occurred in the n experiments. Then the law of large numbers says that
— -> P(A) as n -> oo. (4.3.2)
n
As the reasoning goes, to learn about the probability of an event A we need only repeat the
random experiment to get a reasonable estimate of the probability’s value, and if we are not satisfied
with our estimate then we may simply repeat the experiment more times all the while confident that
with more and more experiments our estimate will stabilize to the true value.
The frequentist approach is good because it is relatively light on assumptions and does not
worry about symmetry or claims of objectivity like the measure-theoretic approach does. It is
perfect for the spinning coin experiment. One drawback to the method is that one can never know
the exact value of a probability, only a long-run approximation. It also does not work well with
80
CHAPTER 4. PROBABILITY
experiments that can not be repeated indefinitely, say, the probability that it will rain today, the
chances that you get will get an A in your Statistics class, or the probability that the world is
destroyed by nuclear war.
This approach was espoused by Richard von Mises in the early twentieth century, and some of
his main ideas were incorporated into the measure theory approach. See
http : //www- hi story ,mcs . st-andrews . ac . uk/Biographies/Mises . html
for more.
4.3.3 The Subjective Approach
The subjective approach interprets probability as the experimenter’s degree of belief that the event
will occur. The estimate of the probability of an event is based on the totality of the individual’s
knowledge at the time. As new information becomes available, the estimate is modified accordingly
to best reflect his/her current knowledge. The method by which the probabilities are updated is
commonly done with Bayes’ Rule, discussed in Section 4.8.
So for the coin toss example, a person may have P(Heads) = 1/2 in the absence of additional
information. But perhaps the observer knows additional information about the coin or the thrower
that would shift the probability in a certain direction. For instance, parlor magicians may be trained
to be quite skilled at tossing coins, and some are so skilled that they may toss a fair coin and get
nothing but Heads, indefinitely. I have see n this. It was similarly claimed in Bringing Down the
House [65] that MIT students were accomplished enough with cards to be able to cut a deck to the
same location, every single time. In such cases, one clearly should use the additional information
to assign P(Heads) away from the symmetry value of 1 /2.
This approach works well in situations that cannot be repeated indefinitely, for example, to
assign your probability that you will get an A in this class, the chances of a devastating nuclear
war, or the likelihood that a cure for the common cold will be discovered.
The roots of subjective probability reach back a long time. See
http : //en . wikipedia . org/wiki/Sub j ective_probability
for a short discussion and links to references about the subjective approach.
4.3.4 Equally Likely Model (ELM)
We have seen several approaches to the assignment of a probability model to a given random exper-
iment and they are very different in their underlying interpretation. But they all cross paths when
it comes to the equally likely model which assigns equal probability to all elementary outcomes of
the experiment.
The ELM appears in the measure theory approach when the experiment boasts symmetry of
some kind. If symmetry guarantees that all outcomes have equal “size”, and if outcomes with
equal “size” should get the same probability, then the ELM is a logical objective choice for the
experimenter. Consider the balanced 6-sided die, the fair coin, or the dart board with equal-sized
wedges.
The ELM appears in the subjective approach when the experimenter resorts to indifference or
ignorance with respect to his/her knowledge of the outcome of the experiment. If the experimenter
4.3. MODEL ASSIGNMENT
81
has no prior knowledge to suggest that (s)he prefer Heads over Tails, then it is reasonable for the
him/her to assign equal subjective probability to both possible outcomes.
The ELM appears in the relative frequency approach as a fascinating fact of Nature: when we
flip balanced coins over and over again, we observe that the proportion of times that the coin comes
up Heads tends to 1/2. Of course if we assume that the measure theory applies then we can prove
that the sample proportion must tend to 1/2 as expected, but that is putting the cart before the horse,
in a manner of speaking.
The ELM is only available when there are finitely many elements in the sample space.
4.3.5 How to do it with R
In the prob package, a probability space is an object of outcomes S and a vector of probabilities
(called “probs”) with entries that correspond to each outcome in S. When S is a data frame, we may
simply add a column called probs to S and we will be finished; the probability space will simply
be a data frame which we may call S. In the case that S is a list, we may combine the outcomes
and probs into a larger list, space; it will have two components: outcomes and probs. The only
requirements we need are for the entries of probs to be nonnegative and sum (probs) to be one.
To accomplish this in R, we may use the probspace function. The general syntax is probspace(x
probs), where x is a sample space of outcomes and probs is a vector (of the same length as the
number of outcomes in x). The specific choice of probs depends on the context of the problem,
and some examples follow to demonstrate some of the more common choices.
Example 4.4. The Equally Likely Model asserts that every outcome of the sample space has the
same probability, thus, if a sample space has n outcomes, then probs would be a vector of length
n with identical entries 1 jn. The quickest way to generate probs is with the rep function. We
will start with the experiment of rolling a die, so that n = 6. We will construct the sample space,
generate the probs vector, and put them together with probspace.
> outcomes <- rolldie(l)
XI
1 1
2 2
3 3
4 4
5 5
6 6
> p <- rep (1/6, times = 6)
[1] 0.1666667 0.1666667 0.1666667 0.1666667 0.1666667 0.1666667
> probspace (outcomes, probs = p)
XI probs
1 1 0.1666667
2 2 0.1666667
3 3 0.1666667
4 4 0.1666667
5 5 0.1666667
6 6 0.1666667
82
CHAPTER 4. PROBABILITY
The probspace function is designed to save us some time in many of the most common situa-
tions. For example, due to the especial simplicity of the sample space in this case, we could have
achieved the same result with only (note the name change for the first column)
> probspace(l:6, probs = p)
x probs
1 1 0.1666667
2 2 0.1666667
3 3 0.1666667
4 4 0.1666667
5 5 0.1666667
6 6 0.1666667
Further, since the equally likely model plays such a fundamental role in the study of probability
the probspace function will assume that the equally model is desired if no probs are specified.
Thus, we get the same answer with only
> probspace (1 : 6)
x probs
1 1 0.1666667
2 2 0.1666667
3 3 0.1666667
4 4 0.1666667
5 5 0.1666667
6 6 0.1666667
And finally, since rolling dice is such a common experiment in probability classes, the rolldie
function has an additional logical argument makespace that will add a column of equally likely
probs to the generated sample space:
> rolldie(l , makespace = TRUE)
XI probs
1 1 0.1666667
2 2 0.1666667
3 3 0.1666667
4 4 0.1666667
5 5 0.1666667
6 6 0.1666667
or just rolldie (1 , TRUE). Many of the other sample space functions (tosscoin, cards, roulette,
etc.) have similar makespace arguments. Check the documentation for details.
One sample space function that does NOT have a makespace option is the urnsamples func-
tion. This was intentional. The reason is that under the varied sampling assumptions the outcomes
in the respective sample spaces are NOT, in general, equally likely. It is important for the user to
carefully consider the experiment to decide whether or not the outcomes are equally likely and then
use probspace to assign the model.
4.4. PROPERTIES OF PROBABILITY
83
Example 4.5. An unbalanced coin. While the makespace argument to tosscoin is useful to
represent the tossing of a fair coin, it is not always appropriate. For example, suppose our coin is
not perfectly balanced, for instance, maybe the “H” side is somewhat heavier such that the chances
of a H appearing in a single toss is 0.70 instead of 0.5. We may set up the probability space with
> probspace(tosscoin(l) , probs = cfffl.7, 6.3))
tossl probs
1 H 8.7
2 T 8.3
The same procedure can be used to represent an unbalanced die, roulette wheel, etc.
4.3.6 Words of Warning
It should be mentioned that while the splendour of R is uncontested, it, like everything else, has
limits both with respect to the sample/probability spaces it can manage and with respect to the finite
accuracy of the representation of most numbers (see the R FAQ 7.31). When playing around with
probability, one may be tempted to set up a probability space for tossing 100 coins or rolling 50
dice in an attempt to answer some scintillating question. (Bear in mind: rolling a die just 9 times
has a sample space with over 10 million outcomes.)
Alas, even if there were enough RAM to barely hold the sample space (and there were enough
time to wait for it to be generated), the infinitesimal probabilities that are associated with so many
outcomes make it difficult for the underlying machinery to handle reliably. In some cases, special
algorithms need to be called just to give something that holds asymptotically. User beware.
4.4 Properties of Probability
4.4.1 Probability Functions
A probability function is a rule that associates with each event A of the sample space a unique
number P(A) = p, called the probability of A. Any probability function P satisfies the following
three Kolmogorov Axioms:
Axiom 4.6. P(A) > 0 for any event A C S.
Axiom 4.7. P(S) = 1.
Axiom 4.8. If the events A \ , A 2, A3. . . are disjoint then
(4.4.1)
and furthermore,
(4.4.2)
84
CHAPTER 4. PROBABILITY
The intuition behind the axioms: first, the probability of an event should never be negative.
And since the sample space contains all possible outcomes, its probability should be one, or 100%.
The final axiom may look intimidating, but it simply means that for a sequence of disjoint events
(in other words, sets that do not overlap), their total probability (measure) should equal the sum of
its parts. For example, the chance of rolling a 1 or a 2 on a die is the chance of rolling a 1 plus the
chance of rolling a 2. The connection to measure theory could not be more clear.
4.4.2 Properties
For any events A and B,
1. P(Ae) = 1 - P(A).
Proof. Since A U Ac = S and A Pi Ac = 0, we have
1 = P(S ) = P(A U Ac) = P(A) + P(Ar).
□
2. P(0) = 0.
Proof. Note that 0 = Sc, and use Property 1. □
3. If A c B , then P(A) < P(B).
Proof. Write 5=AU(5fl Ae), and notice that A Pi (B n Ac) = 0; thus
P(B) = P(A U (B n Ac)) = P(A) + P (B n Ac) > P(A),
since P (B n Ac) >0. □
4. 0 < P(A) < 1.
Proof. The left inequality is immediate from Axiom 4.6, and the second inequality follows
from Property 3 since A c S . □
5. The General Addition Rule.
P (A U B) = P(A) + P(B) - P(A n B).
(4.4.3)
More generally, for events Ai, Ai, A3,. . . , A„,
n \ n n-l n
p
l J A, =2 P(A,) - 2 2 F(Ai n Aj) + ■■■+(- ir 1 P f}A,
i= 1 j=i+ 1
v i~\ y
i= 1
(4.4.4)
V 7=1
6. The Theorem of Total Probability. Let If , IT, . . . , Bn be mutually exclusive and exhaustive.
Then
P(A) = P(A n BO + P(A n b2) + ■ ■ ■ + P (A n Bn).
(4.4.5)
4.4. PROPERTIES OF PROBABILITY
85
4.4.3 Assigning Probabilities
A model of particular interest is the equally likely model. The idea is to divide the sample space S
into a finite collection of elementary events {a\ . ai, ■ ■ . , a.v} that are equally likely in the sense that
each a, has equal chances of occurring. The probability function associated with this model must
satisfy P(S) = 1, by Axiom 2. On the other hand, it must also satisfy
N
P(S) = P({«1, <22, • ■ ■ ,<2v}) = P(<2l U <22 U • ■ ■ U <2;v) = ^ P(<2,),
i= 1
by Axiom 3. Since P(a,) is the same for all i, each one necessarily equals 1/A.
For an event A c S , we write it as a collection of elementary outcomes: if A — {a,-, , a,-2, . . . , a,, )
then A has k elements and
P(A) = P(<2„ ) + P (ah) + ■ ■ ■ + P(<2lt),
1 1 1
— + + • * * + ,
N N N
_ k _ #(A)
“ N ~ MS)'
In other words, under the equally likely model, the probability of an event A is determined by the
number of elementary events that A contains.
Example 4.9. Consider the random experiment E of tossing a coin. Then the sample space is
S = {77, T\, and under the equally likely model, these two outcomes have P(77) = P(7) = 1/2.
This model is taken when it is reasonable to assume that the coin is fair.
Example 4.10. Suppose the experiment E consists of tossing a fair coin twice. The sample space
may be represented by S = {HH, HT , 77/, 7 7} . Given that the coin is fair and that the coin is
tossed in an independent and identical manner, it is reasonable to apply the equally likely model.
What is P(at least 1 Head)? Looking at the sample space we see the elements HH, HT, and
T H have at least one Head; thus, P(at least 1 Head) = 3/4.
What is P(no Heads)? Notice that the event {no Heads) = {at least one Head}c, which by Prop-
erty 1 means P(no Heads) = 1 - P(at least one Head) = 1 - 3/4 = 1/4. It is obvious in this simple
example that the only outcome with no Heads is TT, however, this complementation trick is useful
in more complicated circumstances.
Example 4.11. Imagine a three child family, each child being either Boy (B) or Girl (G). An
example sequence of siblings would be BGB. The sample space may be written
f BBB, BGB, GBB, GGB, )
6 " \ BBG, BGG, GBG, GGG J '
Note that for many reasons (for instance, it turns out that girls are slightly more likely to be born
than boys), this sample space is not equally likely. For the sake of argument, however, we will
assume that the elementary outcomes each have probability 1/8.
What is P(exactly 2 Boys)? Inspecting the sample space reveals three outcomes with exactly
two boys: {BBG, BGB, GBB). Therefore P(exactly 2 Boys) = 3/8.
86
CHAPTER 4. PROBABILITY
What is P(at most 2 Boys)? One way to solve the problem would be to count the outcomes
that have 2 or less Boys, but a quicker way would be to recognize that the only way that the event
{at most 2 Boys} does not occur is the event {all Girls}. Thus
P(at most 2 Boys) = 1 - P(GGG) = 1 — 1/8 = 7/8.
Example 4.12. Consider the experiment of rolling a six-sided die, and let the outcome be the face
showing up when the die comes to rest. Then S = {1,2, 3, 4,5, 6). It is usually reasonable to
suppose that the die is fair, so that the six outcomes are equally likely.
Example 4.13. Consider a standard deck of 52 cards. These are usually labeled with the four suits:
Clubs, Diamonds, Hearts, and Spades, and the 13 ranks: 2, 3, 4, ... , 10, Jack (J), Queen (Q), King
(K), and Ace (A). Depending on the game played, the Ace may be ranked below 2 or above King.
Let the random experiment E consist of drawing exactly one card from a well-shuffled deck, and
let the outcome be the face of the card. Define the events A = {draw an Ace) and B = {draw a Club).
Bear in mind: we are only drawing one card.
Immediately we have P(A) = 4/52 since there are four Aces in the deck; similarly, there are 13
Clubs which implies P(B) = 13/52.
What is P(A n B)1 We realize that there is only one card of the 52 which is an Ace and a Club
at the same time, namely, the Ace of Clubs. Therefore P(A C\ B) - 1/52.
To find P(A U B) we may use the above with the General Addition Rule to get
P (A U B) = P(A) + P (B) - P(A n B),
= 4/52+ 13/52- 1/52,
= 16/52.
Example 4.14. Staying with the deck of cards, let another random experiment be the selec-
tion of a five card stud poker hand, where “five card stud” means that we draw exactly five
cards from the deck without replacement, no more, and no less. It turns out that the sample
space S is so large and complicated that we will be obliged to settle for the trivial description
S - {all possible 5 card hands} for the time being. We will have a more precise description later.
What is P(Royal Flush), or in other words, P(A, K, Q, J, 10 all in the same suit)?
It should be clear that there are only four possible royal flushes. Thus, if we could only count
the number of outcomes in S then we could simply divide four by that number and we would have
our answer under the equally likely model. This is the subject of Section 4.5.
4.4.4 How to do it with R
Probabilities are calculated in the prob package with the prob function.
Consider the experiment of drawing a card from a standard deck of playing cards. Let’s denote
the probability space associated with the experiment as S, and let the subsets A and B be defined by
the following:
> S <- cards (makespace = TRUE)
> A <- subset (S, suit == "Heart")
> B <- subset (S, rank %in% 7:9)
Now it is easy to calculate
4.5. COUNTING METHODS
87
> prob(A)
[1] 0.25
Note that we can get the same answer with
> prob(S, suit == "Heart")
[1] 0.25
We also find prob(B)= 0.23 (listed here approximately, but 12/52 actually) and prob(S)= 1.
Internally, the prob function operates by summing the probs column of its argument. It will find
subsets on-the-fly if desired.
We have as yet glossed over the details. More specifically, prob has three arguments: x, which
is a probability space (or a subset of one), event, which is a logical expression used to define a
subset, and given, which is described in Section 4.6.
WARNING. The event argument is used to define a subset of x, that is, the only outcomes used
in the probability calculation will be those that are elements of x and satisfy event simultaneously.
In other words, prob(x, event) calculates
prob (intersect (x , subset(x, event)))
Consequently, x should be the entire probability space in the case that event is non-null.
4.5 Counting Methods
The equally-likely model is a convenient and popular way to analyze random experiments. And
when the equally likely model applies, finding the probability of an event A amounts to nothing
more than counting the number of outcomes that A contains (together with the number of events in
S). Hence, to be a master of probability one must be skilled at counting outcomes in events of all
kinds.
Proposition 4.15. The Multiplication Principle. Suppose that an experiment is composed of two
successive steps. Further suppose that the first step may be performed in n \ distinct ways while the
second step may be performed in «2 distinct ways. Then the experiment may be performed in n
distinct ways.
More generally, if the experiment is composed of k successive steps which may be performed
in n\, n 2, ■ ■ ■ , tik distinct ways, respectively, then the experiment may be performed in npi^ ■ ■ ■ Hk
distinct ways.
Example 4.16. We would like to order a pizza. It will be sure to have cheese (and marinara sauce),
but we may elect to add one or more of the following five (5) available toppings:
pepperoni, sausage, anchovies, olives, and green peppers.
How many distinct pizzas are possible?
There are many ways to approach the problem, but the quickest avenue employs the Multipli-
cation Principle directly. We will separate the action of ordering the pizza into a series of stages.
At the first stage, we will decide whether or not to include pepperoni on the pizza (two possibili-
ties). At the next stage, we will decide whether or not to include sausage on the pizza (again, two
CHAPTER 4. PROBABILITY
possibilities). We will continue in this fashion until at last we will decide whether or not to include
green peppers on the pizza.
At each stage we will have had two options, or ways, to select a pizza to be made. The Multi-
plication Principle says that we should multiply the 2’s to find the total number of possible pizzas:
2 • 2 • 2 • 2 • 2 = 25 = 32.
Example 4.17. We would like to buy a desktop computer to study statistics. We go to a website
to build our computer our way. Given a line of products we have many options to customize our
computer. In particular, there are 2 choices for a processor, 3 different operating systems, 4 levels
of memory, 4 hard drives of differing sizes, and 10 choices for a monitor. How many possible types
of computer must Gell be prepared to build? Answer: 2 • 3 • 4 • 4 • 10 = 960.
4.5.1 Ordered Samples
Imagine a bag with n distinguishable balls inside. Now shake up the bag and select k balls at
random. How many possible sequences might we observe?
Proposition 4.18. The number of ways in which one may select an ordered sample of k subjects
from a population that has n distinguishable members is
• nk if sampling is done with replacement,
• n(n — 1 )(n — 2) • • • (n — k + 1) if sampling is done without replacement.
Recall from calculus the notation for factorials:
1! = 1,
2! = 2-1=2,
3! = 3-2-1 =6,
n\ = n{n - \){n - 2) • • • 3 • 2 • 1.
Fact 4.19. The number of permutations ofn elements is n\.
Example 4.20. Take a coin and flip it 7 times. How many sequences of Heads and Tails are
possible? Answer: 27 = 128.
Example 4.21. In a class of 20 students, we randomly select a class president, a class vice-
president, and a treasurer. How many ways can this be done? Answer: 20 • 19 • 18 = 6840.
Example 4.22. We rent five movies to watch over the span of two nights. We wish to watch 3
movies on the first night. How many distinct sequences of 3 movies could we possibly watch?
Answer: 5 • 4 • 3 = 60.
4.5.2 Unordered Samples
The number of ways in which one may select an unordered sample of k subjects from a population
that has n distinguishable members is
4.5. COUNTING METHODS
89
ordered = TRUE ordered = FALSE
replace = TRUE
replace = FALSE
n\
( n—k ) !
( n-\+k)\
(n-l)\k\
Table 4.1: Sampling k from n objects with urnsamples
• (n — 1 + k)\/[(n - 1) !Ar!] if sampling is done with replacement,
• n\/[kl(n - £)!] if sampling is done without replacement.
The quantity n\/[k\(n - A:)!] is called a binomial coefficient and plays a special role in mathematics;
it is denoted
h\ n!
k) ~ k\(n - k)\
(4.5.1)
and is read “n choose k”.
Example 4.23. You rent five movies to watch over the span of two nights, but only wish to watch
3 movies the first night. Your friend, Fred, wishes to borrow some movies to watch at his house
on the first night. You owe Fred a favor, and allow him to select 2 movies from the set of 5. How
many choices does Fred have? Answer: Q) =10.
Example 4.24. Place 3 six-sided dice into a cup. Next, shake the cup well and pour out the dice.
How many distinct rolls are possible? Answer: (6 - 1 + 3)!/[(6 - 1 ) !3 !] = = 56.
4.5.3 How to do it with R
The factorial n\ is computed with the command factorial (n) and the binomial coefficient (")
with the command choose (n , k) .
The sample spaces we have computed so far have been relatively small, and we can visually
study them without much trouble. However, it is very easy to generate sample spaces that are
prohibitively large. And while R is wonderful and powerful and does almost everything except
wash windows, even R has limits of which we should be mindful.
But we often do not need to actually generate the sample space; it suffices to count the number
of outcomes. The nsamp function will calculate the number of rows in a sample space made by
urnsamples without actually devoting the memory resources necessary to generate the space.
The arguments are n, the number of (distinguishable) objects in the urn, k, the sample size, and
replace, ordered, as above.
Example 4.25. We will compute the number of outcomes for each of the four urnsamples exam-
ples that we saw in Example 4.2. Recall that we took a sample of size two from an urn with three
distinguishable elements.
90
CHAPTER 4. PROBABILITY
> nsamp(n -3, k - 2,
[1] 9
> nsamp(n = 3, k = 2,
[1] 6
> nsamp(n -3, k - 2,
[1] 3
> nsampCn -3, k - 2,
[1] 6
replace = TRUE, ordered = TRUE)
replace = FALSE, ordered = TRUE)
replace = FALSE, ordered = FALSE)
replace = TRUE, ordered = FALSE)
Compare these answers with the length of the data frames generated above.
The Multiplication Principle
A benefit of nsamp is that it is vectorized so that entering vectors instead of numbers for n, k,
replace, and ordered results in a vector of corresponding answers. This becomes particularly
convenient for combinatorics problems.
Example 4.26. There are 1 1 artists who each submit a portfolio containing 7 paintings for compe-
tition in an art exhibition. Unfortunately, the gallery director only has space in the winners’ section
to accommodate 12 paintings in a row equally spread over three consecutive walls. The director
decides to give the first, second, and third place winners each a wall to display the work of their
choice. The walls boast 31 separate lighting options apiece. How many displays are possible?
Answer: The judges will pick 3 (ranked) winners out of 1 1 (with rep = FALSE, ord = TRUE).
Each artist will select 4 of his/her paintings from 7 for display in a row (rep = FALSE, ord =
TRUE), and lastly, each of the 3 walls has 31 lighting possibilities (rep = TRUE, ord = TRUE).
These three numbers can be calculated quickly with
> n <- c(ll, 7, 31)
> k <- c(3, 4, 3)
> r <- c (FALSE, FALSE, TRUE)
> x <- nsamp (n, k, rep = r, ord = TRUE)
[1] 990 840 29791
(Notice that ordered is always TRUE; nsamp will recycle ordered and replace to the appro-
priate length.) By the Multiplication Principle, the number of ways to complete the experiment is
the product of the entries of x:
> prod(x)
[1] 24774195600
Compare this with the some other ways to compute the same thing:
> (11 * 1® * 9) * (7 ■■■ 6 * 5 * 4) * 313
4.5. COUNTING METHODS
91
[1] 26029080®
or alternatively
> prod(9: 11) * prod(4:7) * 313
[1] 260290800
or even
> prod(factorial(c(ll , 7))/factorial(c(8 , 3))) * 313
[1] 260290800
As one can guess, in many of the standard counting problems there aren’t substantial savings
in the amount of typing; it is about the same using nsarap versus factorial and choose. But the
virtue of nsamp lies in its collecting the relevant counting formulas in a one-stop shop. Ultimately,
it is up to the user to choose the method that works best for him/herself.
Example 4.27. The Birthday Problem. Suppose that there are n people together in a room. Each
person announces the date of his/her birthday in turn. The question is: what is the probability of
at least one match? If we let the event A represent {there is at least one match), then would like to
know P(A), but as we will see, it is more convenient to calculate P(AC).
For starters we will ignore leap years and assume that there are only 365 days in a year. Second,
we will assume that births are equally distributed over the course of a year (which is not true due to
all sorts of complications such as hospital delivery schedules). See http://en.wikipedia.org/
wiki/Birthday_problem for more.
Let us next think about the sample space. There are 365 possibilities for the first person’s
birthday, 365 possibilities for the second, and so forth. The total number of possible birthday
sequences is therefore #(S) = 365".
Now we will use the complementation trick we saw in Example 4.11. We realize that the only
situation in which A does not occur is if there are no matches among all people in the room, that is,
only when everybody’s birthday is different, so
#(AC)
P(A) = 1 - P(Ae) = 1 - —
#(S)
since the outcomes are equally likely. Let us then suppose that there are no matches. The first
person has one of 365 possible birthdays. The second person must not match the first, thus, the
second person has only 364 available birthdays from which to choose. Similarly, the third person
has only 363 possible birthdays, and so forth, until we reach the nth person, who has only 365 —n + 1
remaining possible days for a birthday. By the Multiplication Principle, we have #(AC) = 365 ■
364 • ■ • (365 — n+ 1), and
P(A) = 1 -
365 ■ 364 • ■ ■ (365 -n + Y)
365"
364 363
365 ' 365
(365 - n + 1)
365
(4.5.2)
As a surprising consequence, consider this: how many people does it take to be in the room so
that the probability of at least one match is at least 0.50? Clearly, if there is only n — 1 person in
the room then the probability of a match is zero, and when there are n = 366 people in the room
92
CHAPTER 4. PROBABILITY
o
H — 1
cd
a)
c:
o
-) — '
w
cd
0
O
O
Number of people in room
Figure 4.5.1: The birthday problem
The horizontal line is at p = 0.50 and the vertical line is at n = 23.
there is a 100% chance of a match (recall that we are ignoring leap years). So how many people
does it take so that there is an equal chance of a match and no match?
When I have asked this question to students, the usual response is “somewhere around n = 180
people” in the room. The reasoning seems to be that in order to get a 50% chance of a match, there
should be 50% of the available days to be occupied. The number of students in a typical classroom
is 25, so as a companion question I ask students to estimate the probability of a match when there
are n = 25 students in the room. Common estimates are a 1%, or 0.5%, or even 0.1% chance of a
match. After they have given their estimates, we go around the room and each student announces
their birthday. More often than not, we observe a match in the class, to the students’ disbelief.
Students are usually surprised to hear that, using the formula above, one needs only n = 23
students to have a greater than 50% chance of at least one match. Figure 4.5.1 shows a graph of the
birthday probabilities:
4.5.4 How to do it with R
We can make the plot in Figure 4.5.1 with the following sequence of commands,
g <- Vectorize (pbirthday . ipsur)
4.6. CONDITIONAL PROBABILITY
93
plot Cl : 5® ,
xlab =
ylab =
main =
abline(h =
abline(v =
g ( 1 : 5®) ,
"Number of people in room",
"ProbCat least one match)",
"The Birthday Problem")
0.5)
23, lty = 2) # dashed line
There is a Birthday problem item in the Probability menu of RcmdrPlugin. IPSUR.
In the base R version, one can compute approximate probabilities for the more general case of
probabilities other than 1/2, for differing total number of days in the year, and even for more than
two matches.
4.6 Conditional Probability
Consider a full deck of 52 standard playing cards. Now select two cards from the deck, in succes-
sion. Let A = {first card drawn is an Ace) and B = {second card drawn is an Ace). Since there are
four Aces in the deck, it is natural to assign P(A) = 4/52. Suppose we look at the first card. What
now is the probability of B ? Of course, the answer depends on the value of the first card. If the first
card is an Ace, then the probability that the second also is an Ace should be 3/51, but if the first
card is not an Ace, then the probability that the second is an Ace should be 4/51. As notation for
these two situations we write
P(B|A) = 3/51, P(B|AC) = 4/51.
Definition 4.28. The conditional probability of B given A, denoted P(B|A), is defined by
V(B\A) - P(^ B), if P(A) > 0. (4.6.1)
We will not be discussing a conditional probability of B given A when P(A) = 0, even though this
theory exists, is well developed, and forms the foundation for the study of stochastic processes3.
Example 4.29. Toss a coin twice. The sample space is given by S — {HH, I IT, TH, 77’}. Let
A = {a head occurs) and B — {a head and tail occur). It should be clear that P(A) = 3/4, P( B) =
2/4, and P(A n B) = 2/4. What now are the probabilities P(A|7?) and Pi B\A)1
P(A|B) =
P(A n B)
PCS)
2/4
2/4
= 1,
in other words, once we know that a Head and Tail occur, we may be certain that
Next
P(B|A) =
P(A n B)
p (A)
2/4 _ 2
3/4 “ 3’
a Head occurs.
which means that given the information that a Head has occurred, we no longer need to account for
the outcome TT, and the remaining three outcomes are equally likely with exactly two outcomes
lying in the set B.
3 Conditional probability in this case is defined by means of conditional expectation , a topic that is well beyond the scope
of this text. The interested reader should consult an advanced text on probability theory, such as Billingsley, Resnick, or
Ash Dooleans-Dade.
94
CHAPTER 4. PROBABILITY
Second Roll
1
2
3
4
5
6
1
X
2
X
o
3
X
o
o
4
0
o
o
5
O
O
0
o
6
O
O
O
o
0
Table 4.2: Rolling two dice
Example 4.30. Toss a six-sided die twice. The sample space consists of all ordered pairs ( i , j )
of the numbers 1,2, . . . , 6, that is, S — {(1, 1), (1,2), . . . , (6, 6)). We know from Section 4.5 that
#(S ) — 62 = 36. Let A = {outcomes match} and B — {sum of outcomes at least 8). The sample
space may be represented by a matrix:
The outcomes lying in the event A are marked with the symbol “X”> the outcomes falling in B
are marked with “O”, and those in both A and B are marked “®”. Now it is clear that P(A) = 6/36,
P(B) = 15/36, and P(A n B) = 3/36. Finally,
P(A|B) =
3/36
15/36
1
5’
P(£|A) =
3/36
6/36
1
2'
Again, we see that given the knowledge that B occurred (the 15 outcomes in the lower right tri-
angle), there are 3 of the 15 that fall into the set A, thus the probability is 3/15. Similarly, given
that A occurred (we are on the diagonal), there are 3 out of 6 outcomes that also fall in B, thus, the
probability of B given A is 1/2.
4.6.1 How to do it with R
Continuing with Example 4.30, the first thing to do is set up the probability space with the rolldie
function.
> library (prob)
> S <- rolldie(2 , makespace = TRUE) # assumes ELM
> head(S) # first few rows
XI X2 probs
1 1 1 0.02777778
2 2 1 0.02777778
3 3 1 0.02777778
4 4 1 0.02777778
5 5 1 0.02777778
6 6 1 0.02777778
Next we define the events
4.6. CONDITIONAL PROBABILITY
95
> A <- subset(S, XI == X2)
> B <- subset(S, XI + X2 >= 8)
And now we are ready to calculate probabilities. To do conditional probability, we use the
given argument of the prob function:
> prob (A, given = B)
[1] 0.2
> prob(B, given = A )
[1] 0.5
Note that we do not actually need to define the events A and B separately as long as we reference
the original probability space S as the first argument of the prob calculation:
> prob(S, X1--X2, given = (XI + X2 >= 8) )
[1] 0.2
> prob(S, X1+X2 >= 8, given = ( X1--X2) )
[1] 0.5
4.6.2 Properties and Rules
The following theorem establishes that conditional probabilities behave just like regular probabili-
ties when the conditioned event is fixed.
Theorem 4.31. For any fixed event A with P(A) > 0,
1. P(B|A) > 0, for all events B c S ,
2. P(S|A) = 1, and
3. IfBh B2, B 3,... are disjoint events, then
U Bk A =J]p(Bt|A).
(4.6.2)
\k= 1
/ k= 1
In other words, P( jA) is a legitimate probability function. With this fact in mind, the following
properties are immediate:
Proposition 4.32. For any events A, B, and C with P(A) > 0,
1 . P(BC|A) = 1 - P(B|A).
2. If B c C then P(B|A) < P(C|A).
3. P[(B U C)\A] = P(B|A) + P(C|A) - P[(B n C|A)].
96
CHAPTER 4. PROBABILITY
4. The Multiplication Rule. For any two events A and B,
P (A n B) = P(A) P(B|A). (4.6.3)
And more generally, for events Ai, A2, A3,. . . , An,
P(A 1 n A2 n • • • n A„) = P(A, ) P(A2|A0 • • • P(A„|Ai n A2 n • • • n A„_i). (4.6.4)
The Multiplication Rule is very important because it allows us to find probabilities in random
experiments that have a sequential structure, as the next example shows.
Example 4.33. At the beginning of the section we drew two cards from a standard playing deck.
Now we may answer our original question, what is Pfboth Aces)?
Pfboth Aces) = P(A n B) = P(A) P(£|A) =
4 3
52 ' 5l
0.00452.
4.6.3 How to do it with R
Continuing Example 4.33, we set up the probability space by way of a three step process. First we
employ the cards function to get a data frame L with two columns: rank and suit. Both columns
are stored internally as factors with 13 and 4 levels, respectively.
Next we sample two cards randomly from the L data frame by way of the urnsamples function.
It returns a list M which contains all possible pairs of rows from L (there are choose (5 2, 2) of
them). The sample space for this experiment is exactly the list M.
At long last we associate a probability model with the sample space. This is right down the
probspace function’s alley. It assumes the equally likely model by default. We call this result N
which is an object of class ps - short for “probability space”.
But do not be intimidated. The object N is nothing more than a list with two elements: outcomes
and probs. The outcomes element is itself just another list, with choose (52 , 2) entries, each one
a data frame with two rows which correspond to the pair of cards chosen. The probs element is
just a vector with choose (52 , 2) entries all the same: 1/choose (52 , 2).
Putting all of this together we do
> library (prob)
> L <- cards O
> M <- urnsamples (L , size = 2)
> N <- probspace (M)
Now that we have the probability space N we are ready to do some probability. We use the prob
function, just like before. The only trick is to specify the event of interest correctly, and recall that
we were interested in P(both Aces). But if the cards are both Aces then the rank of both cards
should be "A", which sounds like a job for the all function:
> prob(N, all (rank == "A"))
[1] 0.8Q4524887
Note that this value matches what we found in Example 4.33, above. We could calculate all
sorts of probabilities at this point; we are limited only by the complexity of the event’s computer
representation.
4.6. CONDITIONAL PROBABILITY
97
Example 4.34. Consider an urn with 10 balls inside, 7 of which are red and 3 of which are green.
Select 3 balls successively from the urn. Let A = {1st ball is red), B = {2nd ball is redj, and C =
{3rd ball is red}. Then
P(all 3 balls are red) = P(A n B n C) = ^ ^ « 0.2917.
4.6.4 How to do it with R
Example 4.34 is similar to Example 4.33, but it is even easier. We need to set up an urn (vector
L) to hold the balls, we sample from L to get the sample space (data frame M), and we associate a
probability vector (column probs) with the outcomes (rows of M) of the sample space. The final
result is a probability space (an ordinary data frame N).
It is easier for us this time because our urn is a vector instead of a cards () data frame. Be-
fore there were two dimensions of information associated with the outcomes (rank and suit) but
presently we have only one dimension (color).
> library (prob)
> L <- rep(c("red" , "green"), times = c(7, 3))
> N <- urnsamples(L , size = 3, replace = FALSE, ordered = TRUE)
> N <- probspace(M)
Now let us think about how to set up the event {all 3 balls are red). Rows of N that satisfy this
condition have Xl=="red"&X2=="red"&X3=="red", but there must be an easier way. Indeed,
there is. The isrep function (short for “is repeated”) in the prob package was written for this
purpose. The command isrep(N, "red" , 3) will test each row of N to see whether the value
"red" appears 3 times. The result is exactly what we need to define an event with the prob
function. Observe
> prob(N, isrepCN, "red" , 3))
[1] 0.2916667
Note that this answer matches what we found in Example 4.34. Now let us try some other
probability questions. What is the probability of getting two "red"s?
> prob(N, isrepCN, "red" , 2))
[1] 0.525
Note that the exact value is 21/40; we will learn a quick way to compute this in Section 5.6.
What is the probability of observing "red", then "green", then "red"?
> prob(N, isin(N, c("red", "green", "red"), ordered = TRUE))
[1] 0.175
Note that the exact value is 7/20 (do it with the Multiplication Rule). What is the probability
of observing "red", "green", and "red", in no particular order?
98
CHAPTER 4. PROBABILITY
> prob(N, isin(N, c("red", "green", "red")))
[1] 0.525
We already knew this. It is the probability of observing two "red"s, above.
Example 4.35. Consider two urns, the brst with 5 red balls and 3 green balls, and the second with
2 red balls and 6 green balls. Your friend randomly selects one ball from the first urn and transfers
it to the second urn, without disclosing the color of the ball. You select one ball from the second
urn. What is the probability that the selected ball is red? Let A - {transferred ball is red) and
B = {selected ball is red). Write
B - S C\ B
= (A U Ac) n B
= (A OB) U (Ac n B)
and notice that A n B and Ac n B are disjoint. Therefore
P(fi) = P(A n B) + P(AC n B)
= P(A) P(B|A) + P(AC) P(B|AC)
_ 5 3 3 2
" 8 ' 9 + 8 ' 9
_ 21
“ 72
(which is 7/24 in lowest terms).
Example 4.36. We saw the RcmdrTestDrive data set in Chapter 2 in which a two-way table of
the smoking status versus the gender was
gender
smoke
Female
Male
Sum
No
80
54
134
Yes
15
19
34
Sum
95
73
168
If one person were selected at random from the data set, then we see from the two-way table
that P(Female) = 70/168 and P(Smoker) = 32/168. Now suppose that one of the subjects quits
smoking, but we do not know the person’s gender. If we select one subject at random, what now is
P(Female)? Let A = {the quitter is a female) and B = {selected person is a female). Write
B = S OB
= (A U Ac) n B
= (A n B) u (Ac n B)
4.7. INDEPENDENT EVENTS
99
and notice that A Pi B and Ac Pi B are disjoint. Therefore
P (B) = P (A n B) + P(AC n B),
= P(A)P(fi|A) + P(Ac)P(B|Ar),
_ 5 3 3 2
~ 8 ' 9 + 8 ' 9’
_ 21
~ 72’
(which is 7/24 in lowest terms).
Using the same reasoning, we can return to the example from the beginning of the section and
show that
P({second card is an Ace}) = 4/52.
4.7 Independent Events
Toss a coin twice. The sample space is S — {HH, HT, TH , TT). We know that P(lst toss is H) =
2/4, P(2nd toss is H) = 2/4, and P(both II) = 1/4. Then
P(2nd toss is H \ 1st toss is H) -
P(both H)
P(lst toss is H)’
= 1/1
2/4’
= P(2nd toss is H).
Intuitively, this means that the information that the first toss is H has no bearing on the probability
that the second toss is H. The coin does not remember the result of the first toss.
Definition 4.37. Events A and B are said to be independent if
P(A Pi B) = P(A) P(B). (4.7.1)
Otherwise, the events are said to be dependent.
The connection with the above example stems from the following. We know from Section 4.6
that when P(Z?) > 0 we may write
P(A|B) =
P(A Pi B)
PCS)
(4.7.2)
In the case that A and B are independent, the numerator of the fraction factors so that P(B) cancels
with the result:
P(A|B) = P(A) when A, B are independent. (4.7.3)
The interpretation in the case of independence is that the information that the event B occurred
does not influence the probability of the event A occurring. Similarly, P(B|A) = P( B). and so
100
CHAPTER 4. PROBABILITY
the occurrence of the event A likewise does not affect the probability of event B. It may seem
more natural to define A and B to be independent when P(A|B) = P(A); however, the conditional
probability P(A|Z?) is only defined when P( B) > 0. Our definition is not limited by this restriction.
It can be shown that when P(A), P(B) > 0 the two notions of independence are equivalent.
Proposition 4.38. If the events A and B are independent then
• A and Bc are independent,
• Ac and B are independent,
• Ac and Bc are independent.
Proof. Suppose that A and B are independent. We will show the second one; the others are similar.
We need to show that
P(AC 0 5) = P(AC) P (B).
To this end, note that the Multiplication Rule, Equation 4.6.3 implies
P(Af O B) = P(fi)P(Ac|fi),
= P(B)[1 -P(A|£)],
= P(£)P(Af).
□
Definition 4.39. The events A, B, and C are mutually independent if the following four conditions
are met:
P(A 0 5) = P(A) P (B),
P (A n C) = P(A) P(C),
P (B n C) = P (B) P(C),
and
P(A050C) = P(A) P (B) P(C).
If only the first three conditions hold then A, B, and C are said to be independent pairwise. Note
that pairwise independence is not the same as mutual independence when the number of events is
larger than two.
We can now deduce the pattern for n events, n > 3. The events will be mutually independent
only if they satisfy the product equality pairwise, then in groups of three, in groups of four, and
so forth, up to all n events at once. For n events, there will be 2" — n — 1 equations that must
be satisfied (see Exercise 4.1). Although these requirements for a set of events to be mutually
independent may seem stringent, the good news is that for most of the situations considered in this
book the conditions will all be met (or at least we will suppose that they are).
Example 4.40. Toss ten coins. What is the probability of observing at least one Head? Answer:
Let A,- = jthe Ith coin shows //} , i = 1,2, . . . , 10. Supposing that we toss the coins in such a way
that they do not interfere with each other, this is one of the situations where all of the A, may be
4.7. INDEPENDENT EVENTS
101
considered mutually independent due to the nature of the tossing. Of course, the only way that
there will not be at least one Head showing is if all tosses are Tails. Therefore,
which is approximately 0.9990234.
4.7.1 How to do it with R
Example 4.41. Toss ten coins. What is the probability of observing at least one Head?
> S <- tosscoin(l<3 , makespace = TRUE )
> A <- subset(S, isrep(S, vals = "T", nrep = 16)))
> 1 - prob(A)
[1] 0.999(3234
Compare this answer to what we got in Example 4.40.
Independent, Repeated Experiments
Generalizing from above it is common to repeat a certain experiment multiple times under identical
conditions and in an independent manner. We have seen many examples of this already: tossing a
coin repeatedly, rolling a die or dice, etc.
The iidspace function was designed specifically for this situation. It has three arguments: x,
which is a vector of outcomes, ntrials, which is an integer telling how many times to repeat the
experiment, and probs to specify the probabilities of the outcomes of x in a single trial.
Example 4.42. An unbalanced coin (continued, see Example 4.5). It was easy enough to set up
the probability space for one unbalanced toss, however, the situation becomes more complicated
when there are many tosses involved. Clearly, the outcome HHH should not have the same proba-
bility as TTT. which should again not have the same probability as HTH. At the same time, there
is symmetry in the experiment in that the coin does not remember the face it shows from toss to
toss, and it is easy enough to toss the coin in a similar way repeatedly.
We may represent tossing our unbalanced coin three times with the following:
> iidspace(c("H" , "T") , ntrials = 3, probs = cfffl.7, <3.3))
XI X2 X3 probs
1 H H H 8.343
2 T H H 8.147
3 H T H 8.147
4 T T H 8.863
5 H H T 8.147
P(at least one H) - 1 - Pfall T),
= 1 - P (A\ n Ac2 n • • • n Aj0),
= l-P(A‘1)P(A^)---P(A^0),
102
CHAPTER 4. PROBABILITY
6 T H T 8.063
7 H T T 0.063
8 T T T 0.027
As expected, the outcome HHH has the largest probability, while TTT has the smallest. (Since
the trials are independent, P (HHH) = 0.7 ’ and W(TTT) = 0.33, etc .) Note that the result of the
function call is a probability space, not a sample space (which we could construct already with the
tosscoin or urnsamples functions). The same procedure could be used to model an unbalanced
die or any other experiment that may be represented with a vector of possible outcomes.
Note that iidspace will assume x has equally likely outcomes if no probs argument is speci-
fied. Also note that the argument x is a vector, not a data frame. Something like iidspace (tosscoin (1) ,
would give an error.
4.8 Bayes’ Rule
We mentioned the subjective view of probability in Section 4.3. In this section we introduce a rule
that allows us to update our probabilities when new information becomes available.
Theorem 4.43. (Bayes’ Rule). Let B\, Bi, . . . , Bn be mutually exclusive and exhaustive and let A
be an event with P(A) > 0. Then
P(B,|A) =
W(Bk)V(A\Bk)
k=\,2,...,n.
(4.8.1)
Proof. The proof follows from looking at P( If. n A) in two different ways. For simplicity, suppose
that P(Bf) > 0 for all k. Then
P(A)P0B,|A) = P(B, n A) = P(Bfc) P(A|Bjt).
Since P(A) > 0 we may divide through to obtain
P(B*|A) =
W(Bk)V(A\Bk)
P (A)
Now remembering that \Bk) is a partition, the Theorem of Total Probability (Equation 4.4.5) gives
the denominator of the last expression to be
P(A) - Yj P(B* n = 2 P(Bi } P(AI^)‘
k=\ k= 1
□
What does it mean? Usually in applications we are given (or know) a priori probabilities PIB^).
We go out and collect some data, which we represent by the event A. We want to know: how do
we update P(Ba) to P (fi/jA)? The answer: Bayes’ Rule.
4.8. BAYES’ RULE
103
Example 4.44. Misfiling Assistants. In this problem, there are three assistants working at a com-
pany: Moe, Larry, and Curly. Their primary job duty is to file paperwork in the filing cabinet when
papers become available. The three assistants have different work schedules:
Moe
Larry
Curly
Workload
60%
30%
10%
That is, Moe works 60% of the time, Larry works 30% of the time, and Curly does the remaining
10%, and they file documents at approximately the same speed. Suppose a person were to select
one of the documents from the cabinet at random. Let M be the event
M = {Moe filed the document)
and let L and C be the events that Larry and Curly, respectively, filed the document. What are
these events’ respective probabilities? In the absence of additional information, reasonable prior
probabilities would just be
Moe
Larry
Curly
Prior Probability
P (M) = 0.60
P(L) = 0.30
P(C) = 0.10
Now, the boss comes in one day, opens up the file cabinet, and selects a file at random. The boss
discovers that the file has been misplaced. The boss is so angry at the mistake that (s)he threatens
to fire the one who erred. The question is: who misplaced the file?
The boss decides to use probability to decide, and walks straight to the workload schedule.
(S)he reasons that, since the three employees work at the same speed, the probability that a ran-
domly selected file would have been filed by each one would be proportional to his workload. The
boss notifies Moe that he has until the end of the day to empty his desk.
But Moe argues in his defense that the boss has ignored additional information. Moe’s like-
lihood of having misfiled a document is smaller than Larry’s and Curly’s, since he is a diligent
worker who pays close attention to his work. Moe admits that he works longer than the others,
but he doesn’t make as many mistakes as they do. Thus, Moe recommends that - before making a
decision - the boss should update the probability (initially based on workload alone) to incorporate
the likelihood of having observed a misfiled document.
And, as it turns out, the boss has information about Moe, Larry, and Curly’s filing accuracy in
the past (due to historical performance evaluations). The performance information may be repre-
sented by the following table:
Moe
Larry
Curly
Misfile Rate
0.003
0.007
0.010
In other words, on the average, Moe misfiles 0.3% of the documents he is supposed to file.
Notice that Moe was correct: he is the most accurate filer, followed by Larry, and lastly Curly. If
the boss were to make a decision based only on the worker’s overall accuracy, then Curly should
get the axe. But Curly hears this and interjects that he only works a short period during the day,
and consequently makes mistakes only very rarely; there is only the tiniest chance that he misfiled
this particular document.
104
CHAPTER 4. PROBABILITY
The boss would like to use this updated information to update the probabilities for the three
assistants, that is, (s)he wants to use the additional likelihood that the document was misfiled to
update his/her beliefs about the likely culprit. Let A be the event that a document is misfiled. What
the boss would like to know are the three probabilities
P(M|A), P(L|A), and P(C|A).
We will show the calculation for P(M|A), the other two cases being similar. We use Bayes’ Rule
in the form
P(MnA)
P(M|A) = —puT'
Let’s try to find P(M n A), which is just P (M) ■ P(A|M) by the Multiplication Rule. We already
know P(M) = 0.6 and P(A|M) is nothing more than Moe’s misfile rate, given above to be P(A|M) =
0.003. Thus, we compute
P (M n A) = (0.6)(0.003) = 0.0018.
Using the same procedure we may calculate
P(L|A) = 0.0021 and P(C|A) = 0.0010.
Now let’s find the denominator, P(A). The key here is the notion that if a hie is misplaced, then
either Moe or Larry or Curly must have hied it; there is no one else around to do the mishling.
Further, these possibilities are mutually exclusive. We may use the Theorem of Total Probability
4.4.5 to write
P(A) = P(A n M) + P(A n L) + P(A fi C).
Luckily, we have computed these above. Thus
P(A) = 0.0018 + 0.0021 + 0.0010 = 0.0049.
Therefore, Bayes’ Rule yields
P(M|A) =
0.0018
0.0049
0.37.
This last quantity is called the posterior probability that Moe misfiled the document, since it incor-
porates the observed data that a randomly selected hie was misplaced (which is governed by the
mishle rate). We can use the same argument to calculate
Moe
Larry
Curly
Posterior Probability P(M|A) ~ 0.37
P(L|A) * 0.43
P(C|A) * 0.20
The conclusion: Larry gets the axe. What is happening is an intricate interplay between the
time on the job and the mishle rate. It is not obvious who the winner (or in this case, loser) will be,
and the statistician needs to consult Bayes’ Rule to determine the best course of action.
Example 4.45. Suppose the boss gets a change of heart and does not hre anybody. But the next
day (s)he randomly selects another hie and again finds it to be misplaced. To decide whom to
hre now, the boss would use the same procedure, with one small change. (S)he would not use
the prior probabilities 60%, 30%, and 10%; those are old news. Instead, she would replace the
prior probabilities with the posterior probabilities just calculated. After the math she will have new
posterior probabilities, updated even more from the day before.
In this way, probabilities found by Bayes’ rule are always on the cutting edge, always updated
with respect to the best information available at the time.
4.8. BAYES’ RULE
105
4.8.1 How to do it with R
There are not any special functions for Bayes’ Rule in the prob package, but problems like the
ones above are easy enough to do by hand.
Example 4.46. Misfiling assistants (continued from Example 4.44). We store the prior probabili-
ties and the likelihoods in vectors and go to town.
> prior <- c(0.6, (9.3, (9 . 1)
> like <- c(79.003, (9.0(97, &.Q1)
> post <- prior * like
> post/sum(post )
[1] 0.3673469 0.4285714 0.2040816
Compare these answers with what we got in Example 4.44. We would replace prior with
post in a future calculation. We could raise like to a power to see how the posterior is affected
by future document mistakes. (Do you see why? Think back to Section 4.7.)
Example 4.47. Let us incorporate the posterior probability (post) information from the last exam-
ple and suppose that the assistants misfile seven more documents. Using Bayes’ Rule, what would
the new posterior probabilities be?
> newprior <- post
> post <- newprior * likeA7
> post/sum(post )
[1] 0.0003355044 0.1473949328 0.8522695627
We see that the individual with the highest probability of having misfiled all eight documents
given the observed data is no longer Larry, but Curly.
There are two important points. First, we did not divide post by the sum of its entries until
the very last step; we do not need to calculate it, and it will save us computing time to postpone
normalization until absolutely necessary, namely, until we finally want to interpret them as proba-
bilities.
Second, the reader might be wondering what the boss would get if (s)he skipped the intermedi-
ate step of calculating the posterior after only one misfiled document. What if she started from the
original prior, then observed eight misfiled documents, and calculated the posterior? What would
she get? It must be the same answer, of course.
> fastpost <- prior * likeA8
> fastpost/sumC fastpost)
[1] 0.0003355044 0.1473949328 0.8522695627
Compare this to what we got in Example 4.45.
106
CHAPTER 4. PROBABILITY
4.9 Random Variables
We already know about experiments, sample spaces, and events. In this section, we are interested
in a number that is associated with the experiment. We conduct a random experiment E and after
learning the outcome to in S we calculate a number X. That is, to each outcome a> in the sample
space we associate a number X(at ) = x.
Definition 4.48. A random variable X is a function X : S — > R that associates to each outcome
at e S exactly one number X(a>) = x.
We usually denote random variables by uppercase letters such as X, Y, and Z, and we denote
their observed values by lowercase letters x, y, and z. Just as S is the set of all possible outcomes
of E, we call the set of all possible values of X the support of X and denote it by S x-
Example 4.49. Let E be the experiment of flipping a coin twice. We have seen that the sample
space is S = {////, I IT, TH, IT). Now define the random variable X = the number of heads.
That is, for example, X(HH ) = 2, while X(HT ) = 1. We may make a table of the possibilities:
(x) 6 S
HH
HT
TH
TT
X(a>) = x
2
1
1
0
Taking a look at the second row of the table, we see that the support of X - the set of all numbers
that X assumes - would be Sx — {0, 1,2).
Example 4.50. Let E be the experiment of flipping a coin repeatedly until observing a Head.
The sample space would be S = {H, TH, TTH, TTTH, . . .}. Now define the random variable
Y = the number of Tails before the first head. Then the support of Y would be S y — {0, 1, 2, . . .}.
Example 4.51. Let E be the experiment of tossing a coin in the air, and define the random variable
Z = the time (in seconds) until the coin hits the ground. In this case, the sample space is inconve-
nient to describe. Yet the support of Z would be (0, oo). Of course, it is reasonable to suppose that
the coin will return to Earth in a short amount of time; in practice, the set (0, oo) is admittedly too
large. However, we will find that in many circumstances it is mathematically convenient to study
the extended set rather than a restricted one.
There are important differences between the supports of X, Y , and Z. The support of A is a
finite collection of elements that can be inspected all at once. And while the support of Y cannot be
exhaustively written down, its elements can nevertheless be listed in a naturally ordered sequence.
Random variables with supports similar to those of X and Y are called discrete random variables.
We study these in Chapter 5.
In contrast, the support of Z is a continuous interval, containing all rational and irrational pos-
itive real numbers. For this reason4, random variables with supports like Z are called continuous
random variables, to be studied in Chapter 6.
4.9.1 How to do it with R
The primary vessel for this task is the addrv function. There are two ways to use it, and we will
describe both.
4This isn’t really the reason, but it serves as an effective litmus test at the introductory level. See Billingsley or Resnick.
4.9. RANDOM VARIABLES
107
Supply a Defining Formula
The first method is based on the transform function. See ?transform. The idea is to write
a formula defining the random variable inside the function, and it will be added as a column to
the data frame. As an example, let us roll a 4-sided die three times, and let us define the random
variable U = XI - X2 + A3.
> S <- rolldie(3 , nsides = 4, makespace = TRUE)
> S <- addrvCS, U = XI - X2 + X3)
Now let’s take a look at the values of U. In the interest of space, we will only reproduce the
first few rows of S (there are 43 = 64 rows in total).
> head(S)
XI X2 X3 U probs
11111 0. 815625
22112 0.015625
33113 0.015625
44114 0.015625
51210 0.015625
62211 0.015625
We see from the U column it is operating just like it should. We can now answer questions like
> prob(S, U > 6)
[1] 0.015625
Supply a Function
Sometimes we have a function laying around that we would like to apply to some of the outcome
variables, but it is unfortunately tedious to write out the formula defining what the new variable
would be. The addrv function has an argument FUN specifically for this case. Its value should be a
legitimate function from R, such as sum, mean, median, etc. Or, you can define your own function.
Continuing the previous example, let’s define V = max(Al , A2, A3) and W — XI + X2 + A3.
> S <- addrvCS, FUN = max, invars = c("Xl", "X2" , "X3"), name = "V")
> S <- addrvCS, FUN = sum, invars = cC'Xl", "X2" , "X3"), name = "W")
> headCS)
XI X2 X3 U V W probs
1 1 1 1113 0.015625
2 2 1 1224 0.015625
3 3 1 1335 0.015625
4 4 1 1446 0.015625
5 1 2 1024 0.015625
6 2 2 1125 0.015625
108
CHAPTER 4. PROBABILITY
Notice that addrv has an invars argument to specify exactly to which columns one would
like to apply the function FUN. If no input variables are specified, then addrv will apply FUN to
all non-probs columns. Further, addrv has an optional argument name to give the new variable;
this can be useful when adding several random variables to a probability space (as above). If not
specified, the default name is “X”.
Marginal Distributions
As we can see above, often after adding a random variable V to a probability space one will find
that V has values that are repeated, so that it becomes difficult to understand what the ultimate
behavior of V actually is. We can use the marginal function to aggregate the rows of the sample
space by values of V, all the while accumulating the probability associated with V’s distinct values.
Continuing our example from above, suppose we would like to focus entirely on the values and
probabilities of V — max(Al , X2, X3).
> marginal (S , vars = "V")
V probs
1 1 0.815625
2 2 0.109375
3 3 0.296875
4 4 0.578125
We could save the probability space of V in a data frame and study it further, if we wish. As
a final remark, we can calculate the marginal distributions of multiple variables desired using the
vars argument. For example, suppose we would like to examine the joint distribution of V and W.
> marginal(S, vars = c("V", "W"))
V
W
probs
1
1
3
0.015625
2
2
4
0.046875
3
2
5
0.046875
4
3
5
0.046875
5
2
6
0.015625
6
3
6
0.093750
7
4
6
0.046875
8
3
7
0.093750
9
4
7
0.093750
10
3
8
0.046875
11
4
8
0.140625
12
3
9
0.015625
13
4
9
0.140625
14
4
10
0.093750
15
4
11
0.046875
16
4
12
0.015625
Note that the default value of vars is the names of all columns except probs. This can be
useful if there are duplicated rows in the probability space.
4.9. RANDOM VARIABLES
109
Chapter Exercises
Exercise 4.1. Prove the assertion given in the text: the number of conditions that the events A i, Ao,
An must satisfy in order to be mutually independent is 2" — n — 1. (Hint: think about Pascal’s
triangle.)
Answer: The events must satisfy the product equalities two at a time, of which there are then
they must satisfy an additional conditions three at a time, and so on, until they satisfy the (”) = 1
condition including all n events. In total, there are
conditions to be satisfied, but the binomial series in the expression on the right is the sum of the
entries of the n* row of Pascal’s triangle, which is 2".
110
CHAPTER 4. PROBABILITY
Chapter 5
Discrete Distributions
In this chapter we introduce discrete random variables, those who take values in a finite or countably
infinite support set. We discuss probability mass functions and some special expectations, namely,
the mean, variance and standard deviation. Some of the more important discrete distributions are
explored in detail, and the more general concept of expectation is defined, which paves the way for
moment generating functions.
We give special attention to the empirical distribution since it plays such a fundamental role
with respect to re sampling and Chapter 13; it will also be needed in Section 10.5.1 where we
discuss the Kolmogorov-Smirnov test. Following this is a section in which we introduce a catalogue
of discrete random variables that can be used to model experiments.
There are some comments on simulation, and we mention transformations of random variables
in the discrete case. The interested reader who would like to learn more about any of the assorted
discrete distributions mentioned here should take a look at Univariate Discrete Distributions by
Johnson et al [50].
What do I want them to know?
• how to choose a reasonable discrete model under a variety of physical circumstances
• the notion of mathematical expectation, how to calculate it, and basic properties
• moment generating functions (yes, I want them to hear about those)
• the general tools of the trade for manipulation of continuous random variables, integration,
etc.
• some details on a couple of discrete models, and exposure to a bunch of other ones
• how to make new discrete random variables from old ones
111
112
CHAPTER 5. DISCRETE DISTRIBUTIONS
5.1 Discrete Random Variables
5.1.1 Probability Mass F unctions
Discrete random variables are characterized by their supports which take the form
Sx — {ni, U2, . . . , Uk) or Sx — {mi , ui, M3 . . .}. (5.1.1)
Every discrete random variable X has associated with it a probability mass function (PMF) fx :
S x — = ► [0, 1] defined by
fx(x) = V(X = x), xeSx. (5.1.2)
Since values of the PMF represent probabilities, we know from Chapter 4 that PMFs enjoy certain
properties. In particular, all PMFs satisfy
1 ■ fx(x) > 0 for x e S ,
2. Sags fx(x) = 1, and
3. PCX' e A) = £ ac. 4 fx(x), for any event A c S .
Example 5.1. Toss a coin 3 times. The sample space would be
S = {HHH, HTH , THH, TTH, HHT, HTT, THT, TTT) .
Now let X be the number of Heads observed. Then X has support S x = {0, 1,2, 3). Assuming that
the coin is fair and was tossed in exactly the same way each time, it is not unreasonable to suppose
that the outcomes in the sample space are all equally likely. What is the PMF of XI Notice that
X is zero exactly when the outcome TTT occurs, and this event has probability 1/8. Therefore,
fx( 0) = 1/8, and the same reasoning shows that fxO) = 1/8. Exactly three outcomes result in
X = 1, thus, fx( 1 ) = 3/8 and fx( 3) holds the remaining 3/8 probability (the total is 1). We can
represent the PMF with a table:
x e Sx
0
1
2
3
Total
fx(x) = P(X = x)
1/8
3/8
3/8
1/8
1
5.1.2 Mean, Variance, and Standard Deviation
There are numbers associated with PMFs. One important example is the mean //, also known as
1E2C (which we will discuss later):
/i = EX = J]x/Kx), (5.1.3)
xeS
provided the (potentially infinite) series 2 Mfx(x) is convergent. Another important number is the
variance:
O-2 = ^(x - nffxix), (5. 1 .4)
xeS
which can be computed (see Exercise 5.3) with the alternate formula <x2 = 2 x2fx(x) - /r. Directly
defined from the variance is the standard deviation <x = Vcr2.
5.1. DISCRETE RANDOM VARIABLES
113
Example 5.2. We will calculate the mean of X in Example 5.1.
We interpret \x = 3.5 by reasoning that if we were to repeat the random experiment many times,
independently each time, observe many corresponding outcomes of the random variable X, and
take the sample mean of the observations, then the calculated value would fall close to 3.5. The
approximation would get better as we observe more and more values of X (another form of the Law
of Large Numbers; see Section 4.3). Another way it is commonly stated is that X is 3.5 “on the
average” or “in the long run”.
Remark 5.3. Note that although we say X is 3.5 on the average, we must keep in mind that our X
never actually equals 3.5 (in fact, it is impossible for X to equal 3.5).
Related to the probability mass function fx(x) = PCX’ = x) is another important function called
the cumulative distribution function (CDF), Fx. It is defined by the formula
We know that all PMFs satisfy certain properties, and a similar statement may be made for
CDFs. In particular, any CDF Fx satisfies
• Fx is nondecreasing {t\ < /t implies Fx(t\) < Fx{t/)).
• Fx is right-continuous (lim,_*a+ Fx(t) = Fx{a) for all a e R).
• lim,^_oo Fx(t) = 0 and limf_)ra Fx(t) = 1.
We say that X has the distribution Fx and we write X ~ Fx. In an abuse of notation we will also
write X ~ fx and for the named distributions the PMF or CDF will be identified by the family name
instead of the defining formula.
5.1.3 How to do it with R
The mean and variance of a discrete random variable is easy to compute at the console. Let’s return
to Example 5.2. We will start by defining a vector x containing the support of X , and a vector £ to
contain the values of fx at the respective outcomes in x:
> x <- cf®, 1,2,3)
> f <- cCl/8, 3/8, 3/8, 1/8)
To calculate the mean /j, we need to multiply the corresponding values of x and f and add
them. This is easily accomplished in R since operations on vectors are performed element-wise
(see Section 2.3.4):
> mu <- sum(x * f)
> mu
Fx(t) = W(X < f), — oo < t < oo.
(5.1.5)
[1] 1.5
114
CHAPTER 5. DISCRETE DISTRIBUTIONS
To compute the variance <x2, we subtract the value of mu from each entry in x, square the
answers, multiply by f, and sum. The standard deviation <j is simply the square root of cr2.
> sigma2 <- sum((x-mu) *2 * f)
> sigma2
[1] 0.75
> sigma <- sqrt(sigma2)
> sigma
[1] 0.8660254
Finally, we may find the values of the CDF Fx on the support by accumulating the probabilities
in fx with the cumsum function.
> F - cumsum(f)
> F
[1] 0.125 0.500 0.875 1.000
As easy as this is, it is even easier to do with the distrEx package [74]. We define a random
variable X as an object, then compute things from the object such as mean, variance, and standard
deviation with the functions E, var, and sd:
> library (distrEx)
> X <- DiscreteDistribution(supp = 6:3, prob = c(l , 3 , 3 , l)/8)
> E(X); var(X); sd(X)
[1] 1.5
[1] 0.75
[1] 0.8660254
5.2 The Discrete Uniform Distribution
We have seen the basic building blocks of discrete distributions and we now study particular models
that statisticians often encounter in the field. Perhaps the most fundamental of all is the discrete
uniform distribution.
A random variable X with the discrete uniform distribution on the integers 1,2, ... ,m has PMF
We write X ~ disunif(m). A random experiment where this distribution occurs is the choice of
an integer at random between 1 and 100, inclusive. Let X be the number chosen. Then X ~
disunif(m = 100) and
fx(x) = — , x = 1, 2, . . . ,m.
(5.2.1)
m
5.2. THE DISCRETE UNIFORM DISTRIB UTION
115
We find a direct formula for the mean of X ~ disun if (in):
m m i i
11 m + 1
xfx(x) =>*• — = —( 1 + 2 + • • • + m) = — — , (5.2.2)
^ m m 2
X=\ X=1
where we have used the famous identity 1+2 + ■ ■ ■ + m = mini + l)/2. That is, if we repeatedly
choose integers at random from 1 to in then, on the average, we expect to get (m + l)/2. To get the
variance we first calculate
y x2fx(x) = - v x2 =
z—j m Z—i
X=l X— 1
1 2 1 + l)(2m +1) (m + l)(2m + 1)
and finally,
^ x2fx(x ) •
2 (m + l)(2m + 1) / m + 1
*=1
»r - 1
12
(5.2.3)
Example 5.4. Roll a die and let X be the upward face showing. Then in = 6, n — 7/2 = 3.5, and
0-2 = (62 - 1)/ 12 = 35/12.
5.2.1 How to do it with R
From the console: One can choose an integer at random with the sample function. The general
syntax to simulate a discrete uniform random variable is sample (x , size , replace = TRUE) .
The argument x identifies the numbers from which to randomly sample. If x is a number, then
sampling is done from 1 to x. The argument size tells how big the sample size should be, and
replace tells whether or not numbers should be replaced in the urn after having been sampled.
The default option is replace = FALSE but for discrete uniforms the sampled values should be
replaced. Some examples follow.
5.2.2 Examples
• To roll a fair die 3000 times, do sample (6 , size = 3880, replace = TRUE).
• To choose 27 random numbers from 30 to 70, do sample (38 : 78 , size = 27, replace
= TRUE).
• To flip a fair coin 1000 times, do sample(c("H" , "T") , size = 1888, replace =
TRUE).
With the R Commander: Follow the sequence Probability > Discrete Distributions > Discrete
Uniform distribution > Simulate Discrete uniform variates. . . .
Suppose we would like to roll a fair die 3000 times. In the Number of samples field we enter
1. Next, we describe what interval of integers to be sampled. Since there are six faces numbered 1
through 6, we set from = 1, we set to = 6, and set by = 1 (to indicate that we travel from 1 to
6 in increments of 1 unit). We will generate a list of 3000 numbers selected from among 1,2,...,
6, and we store the results of the simulation. For the time being, we select New Data set. Click
OK.
116
CHAPTER 5. DISCRETE DISTRIBUTIONS
Since we are defining a new data set, the R Commander requests a name for the data set. The
default name is Simsetl, although in principle you could name it whatever you like (according to
R’s rules for object names). We wish to have a list that is 3000 long, so we set Sample Size =
388® and click OK.
In the R Console window, the R Commander should tell you that Simsetl has been initialized,
and it should also alert you that There was 1 discrete uniform variate sample stored
in Simset 1 . . To take a look at the rolls of the die, we click View data set and a window opens.
The default name for the variable is disunif . siml.
5.3 The Binomial Distribution
The binomial distribution is based on a Bernoulli trial , which is a random experiment in which
there are only two possible outcomes: success ( S ) and failure (F). We conduct the Bernoulli trial
and let
{1 if the outcome is S ,
0 if the outcome is F.
(5.3.1)
If the probability of success is p then the probability of failure must be 1 - p = q and the PMF of
Xis
fx(x) = px(] - p)] x, x = 0, 1.
(5.3.2)
It is easy to calculate p -1EX = p and IE X2 = p so that cr2 - p - p2 = p( 1 - p).
5.3.1 The Binomial Model
The Binomial model has three defining properties:
• Bernoulli trials are conducted n times,
• the trials are independent,
• the probability of success p does not change between trials.
If X counts the number of successes in the n independent trials, then the PMF of X is
fx(x) = ('')//(! - pf~x, * = 0,1,2 (5.3.3)
We say that X has a binomial distribution and we write X ~ binom(size = n, prob = p). It is
clear that fx(x) > 0 for all x in the support because the value is the product of nonnegative numbers.
We next check that £ f(x) = 1 :
g (")/>*( 1 - pf~x = Ip + (1 - P)T = 1" = 1.
We next find the mean:
5.3. THE BINOMIAL DISTRIBUTION
117
V in - 1)!
in ~ D!
—n ■ P >
1 Zj (x -!)!(„-
( x - !)!(« - x)\
Ip^Cl - p)(»-iH^i)
-np.
A similar argument shows that E X{X - 1) = n(n - 1 )p2 (see Exercise 5.4). Therefore
Example 5.5. A four-child family. Each child may be either a boy (B) or a girl (G). For simplicity
we suppose that IP (71) = P(G) = 1/2 and that the genders of the children are determined indepen-
dently. If we let X count the number of B’s, then X ~ binom(size = 4, prob = 1/2). Further,
P(X = 2) is
The mean number of boys is 4( 1 /2) = 2 and the variance of X is 4( 1 /2)(1 /2) = 1 .
5.3.2 How to do it with R
The corresponding R function for the PMF and CDF are dbinom and pbinom, respectively. We
demonstrate their use in the following examples.
Example 5.6. We can calculate it in R Commander under the Binomial Distribution menu with the
Binomial probabilities menu item.
Pr
0 0.0625
1 0.2500
2 0.3750
3 0.2500
4 0.0625
We know that the binom(size = 4, prob = 1/2) distribution is supported on the integers 0, 1,
2, 3, and 4; thus the table is complete. We can read off the answer to be ¥(X -2) - 0.3750.
cr2 =EX(X- 1) + E A - [EX]2,
=n(n - 1 )p2 + np - (np)2,
—np —np + np — n p ,
-np - np 2 = np( 1 - p).
118
CHAPTER 5. DISCRETE DISTRIBUTIONS
Example 5.7. Roll 12 dice simultaneously, and let X denote the number of 6’s that appear. We
wish to find the probability of getting seven, eight, or nine 6’s. If we let S = {get a 6 on one roll),
then IP(.S’ ) = 1/6 and the rolls constitute Bernoulli trials; thus X ~ binom(size =12, prob = 1/6 )
and our task is to find P(7 < X < 9). This is just
Again, one method to solve this problem would be to generate a probability mass table and add
up the relevant rows. However, an alternative method is to notice that P(7 < X < 9) = P(A <
9) - P(A < 6) = Fx( 9) - Fx( 6), so we could get the same answer by using the Binomial tail
probabilities. . . menu in the R Commander or the following from the command line:
> pbinom(9, size-12, prob=l/6) - pbinom(6, size-12, prob-1/6)
[1] ® . 001291758
> diff(pbinom(c(6,9) , size = 12, prob - 1/6)) # same thing
[1] 0.001291758
Example 5.8. Toss a coin three times and let X be the number of Heads observed. We know from
before that X ~ binom(size = 3, prob = 1/2) which implies the following PMF:
x = #of Heads
0
1
2
3
f(x) = P(A = x)
1/8
3/8
3/8
1/8
Our next goal is to write down the CDF of X explicitly. The first case is easy: it is impossible
for X to be negative, so if x < 0 then we should have P(A < x) = 0. Now choose a value x
satisfying 0 < x < 1, say, x = 0.3. The only way that X < x could happen would be if A = 0,
therefore, P(A < x) should equal P(A = 0), and the same is true for any 0 < x < 1. Similarly, for
any 1 < x < 2, say, x = 1.73, the event {A < x} is exactly the event {A = 0 or A = 1). Consequently,
P(A < x) should equal P(A = 0 or A = 1) = P(A = 0) + P(A = 1). Continuing in this fashion,
we may figure out the values of Fx(x) for all possible inputs — oo < x < °o, and we may summarize
our observations with the following piecewise defined function:
In particular, the CDF of A is defined for the entire real line, R. The CDF is right continuous and
nondecreasing. A graph of the binom(size = 3, prob = 1/2) CDF is shown in Figure 5.3.1.
Example 5.9. Another way to do Example 5.8 is with the distr family of packages [74]. They
use an object oriented approach to random variables, that is, a random variable is stored in an object
0,
x <0,
0 < x < 1,
f, 1 < x < 2,
|, 2 < x < 3,
x >3.
8
Fx(x) = P(A < x) = U + |
4 + 3
8 ' 8
1,
cumulative probability
5.3. THE BINOMIAL DISTRIBUTION
119
Figure 5.3.1: Graph of the binom(size = 3, prob = 1/2) CDF
120
CHAPTER 5. DISCRETE DISTRIBUTIONS
X, and then questions about the random variable translate to functions on and involving X. Random
variables with distributions from the base package are specified by capitalizing the name of the
distribution.
> library(distr)
> X <- Binom(size = 3, prob = 1/2)
> X
Distribution Object of Class: Binom
size: 3
prob : 8.5
The analogue of the dbinom function for X is the d(X) function, and the analogue of the pbinom
function is the p(X) function. Compare the following:
> d(X)Cl) # ptnf of X evaluated at x = 1
[1] 8.375
> p(X)(2) # cdf of X evaluated at x - 2
[1] 8.875
Random variables defined via the distr package may be plotted, which will return graphs of
the PMF, CDF, and quantile function (introduced in Section 6.3.1). See Figure 5.3.2 for an example.
5.3. THE BINOMIAL DISTRIBUTION
121
Given X ~ dbinom(size = n, prob = p).
How to do: with stats (default) with distr
PMF: P(X = x ) dbinom(x, size = n, prob = p) d(X)(x)
CDF: P(X < x ) pbinom(x, size = n, prob = p) p(X)(x)
Simulate k variates rbinomdc, size = n. prob = p) r(X)(k)
For distr need X <- Binom(size =n, prob =p)
Table 5.1: Correspondence between stats and distr
Probability function of Binom(3,
CDF of Binom(3, 0.5)
O
CO
d
CD
d
'sf
d
C\J
d
q
d
-10 1 2 3 4
Quantile function of Binom(3, 0
X
q
p
Figure 5.3.2: The binom(size = 3, prob = 0.5) distribution from the distr package
122
CHAPTER 5. DISCRETE DISTRIBUTIONS
5.4 Expectation and Moment Generating Functions
5.4.1 The Expectation Operator
We next generalize some of the concepts from Section 5.1.2. There we saw that every1 PMF has
two important numbers associated with it:
P = ^ xfx(x), cr2 = ^(x - p)2fx(x). (5.4.1)
xeS xeS
Intuitively, for repeated observations of X we would expect the sample mean to closely approximate
p as the sample size increases without bound. For this reason we call p the expected value of X and
we write p = EX, where IE is an expectation operator.
Definition 5.10. More generally, given a function g we define the expected value of g(X) by
E g(X) = YJg(x)fx(x), (5.4.2)
xeS
provided the (potentially infinite) series £ r \g(x)\f(x) is convergent. We say that E g( X ) exists.
In this notation the variance is cr2 = E(X - p)2 and we prove the identity
E(X -p)2 = EX2 - (E X)2 (5.4.3)
in Exercise 5.3. Intuitively, for repeated observations of X we would expect the sample mean of
the g(X) values to closely approximate E g(X) as the sample size increases without bound.
Let us take the analogy further. If we expect g(X) to be close to E g(X) on the average, where
would we expect 3 g(X) to be on the average? It could only be 3 Eg(X). The following theorem
makes this idea precise.
Proposition 5.11. For any functions g and h, any random variable X, and any constant c:
1. E c = c,
2. E[c-g(X)] = cEg(X)
3. E[g(X) + h(X)] = E g(X) + E h(X),
provided E g(X) and E h(X) exist.
Proof. Go directly from the definition. For example,
E[c- *(*)] = £ c ' 8(x)fx (x) — c ■ §(x)fx(x) — c~Eig(X).
xeS xeS
□
JNot every, only those PMFs for which the (potentially infinite) series converges.
5.4. EXPECTATION AND MOMENT GENERATING FUNCTIONS
123
5.4.2 Moment Generating Functions
Definition 5.12. Given a random variable X , its moment generating function (abbreviated MGF) is
defined by the formula
Mx(t) = E e'x = t'xfx (x), (5.4.4)
xeS
provided the (potentially infinite) series is convergent for all t in a neighborhood of zero (that is,
for all —e < t < e, for some e > 0).
Note that for any MGF Mx,
Mx( 0) = Ee“ = E 1 = 1. (5.4.5)
We will calculate the MGF for the two distributions introduced above.
Example 5.13. Find the MGF for A ~ disunif(m).
Since f(x) = 1 /m. the MGF takes the form
m 1 i
M(t) = V etx — = — (e' + e2t + ■ ■ ■ + emf), for any t.
m m
X-\
Example 5.14. Find the MGF iotX ~ binom(size = n, prob = p).
Mx(t) = ye‘x{n)p-\\-Pr\
A=0
=(pe' + q)n, for any t.
Applications
We will discuss three applications of moment generating functions in this book. The first is the fact
that an MGF may be used to accurately identify the probability distribution that generated it, which
rests on the following:
Theorem 5.15. The moment generating function, if it exists in a neighborhood of zero, determines
a probability distribution uniquely.
Proof. Unfortunately, the proof of such a theorem is beyond the scope of a text like this one.
Interested readers could consult Billingsley [8]. □
We will see an example of Theorem 5.15 in action.
Example 5.16. Suppose we encounter a random variable which has MGF
Mx(t) = (0.3 + 0.7e')13.
ThenX ~ binom(size = 13, prob = 0.7).
An MGF is also known as a “Laplace Transform” and is manipulated in that context in many
branches of science and engineering.
124
CHAPTER 5. DISCRETE DISTRIBUTIONS
Why is it called a Moment Generating Function?
This brings us to the second powerful application of MGFs. Many of the models we study have a
simple MGF, indeed, which permits us to determine the mean, variance, and even higher moments
very quickly. Let us see why. We already know that
M(0= JV-7(*).
xeS
Take the derivative with respect to t to get
m) = -
^ e'VG) = Yj ^ (e?J70)) = J] xe'xfW’
\xeS / xeS xeS
and so if we plug in zero for t we see
M '(0) = ^ xe°f(x) = Yj xf(x) = P = EX.
xeS xeS
(5.4.6)
(5.4.7)
Similarly, M"(f) = 2 x2&txf(x) so that M"( 0) = E X2. And in general, we can see2 that
M^\ 0) = EXr = rthmoment of Xabout the origin. (5.4.8)
These are also known as raw moments and are sometimes denoted j/r. In addition to these are the
so called central moments fir dehned by
Hr = E{X-ti)r, r = 1,2,... (5.4.9)
Example 5.17. Let X ~ binom(size = n, prob = p) with M(t) — (q + pe!)n. We calculated the
mean and variance of a binomial random variable in Section 5.3 by means of the binomial series.
But look how quickly we find the mean and variance with the moment generating function.
M'{t) =n(q + pe!)n 1 pe' |f=0 ,
=n ■ 1 "~lp,
—np.
And
M"( 0) =n(n - 1 )[q + pel]n~2(pe‘)2 + n[q + pef]”_1pef |r=0 ,
El2 =n(n - l)p2 + np.
Therefore
cr2 = EX2-(EX)2,
=n(n - \)p2 + np - n2p2,
-np - np 2 = npq.
See how much easier that was?
2We are glossing over some significant mathematical details in our derivation. Suffice it to say that when the MGF exists
in a neighborhood of t = 0, the exchange of differentiation and summation is valid in that neighborhood, and our remarks
hold true.
5.5. THE EMPIRICAL DISTRIBUTION
125
Remark 5.18. We learned in this section that M(r)( 0) = E Xr. We remember from Calculus II that
certain functions / can be represented by a Taylor series expansion about a point a, which takes the
form
■A flr>(a)
f(x) = ^ -^(x-aY, for all \x - a\ < R, (5.4.10)
r= 0 r'
where R is called the radius of convergence of the series (see Appendix E.3). We combine the two
to say that if an MGF exists for all t in the interval (— e, e), then we can write
Mx(t) = 2] — —f, for all |f | < e. (5.4.11)
r= 0 r'
5.4.3 How to do it with R
The distrEx package provides an expectation operator E which can be used on random variables
that have been defined in the ordinary distr sense:
> X <- Binom(size = 3, prob = 61.45)
> library (distrEx)
> E(X)
[1] 1.35
> E(3 * X + 4)
[1] 8.85
For discrete random variables with finite support, the expectation is simply computed with
direct summation. In the case that the random variable has infinite support and the function is
crazy, then the expectation is not computed directly, rather, it is estimated by first generating a
random sample from the underlying model and next computing a sample mean of the function of
interest.
There are methods for other population parameters:
> var(X)
[1] 8.7425
> sd(X)
[1] 8.8616844
There are even methods for IQR, mad, skewness, and kurtosis.
5.5 The Empirical Distribution
Do an experiment n times and observe 11 values X\ , X2, x„ of a random variable X. For simplicity
in most of the discussion that follows it will be convenient to imagine that the observed values are
distinct, but the remarks are valid even when the observed values are repeated.
126
CHAPTER 5. DISCRETE DISTRIBUTIONS
Definition 5.19. The empirical cumulative distribution function Fn (written ECDF) is the probabil-
ity distribution that places probability mass 1/n on each of the values x\, %2, • • • , x„. The empirical
PMF takes the form
fx(x) = -, x e {xi,x2, —,xn} . (5.5.1)
n
If the value x, is repeated k times, the mass at x, is accumulated to k/n.
The mean of the empirical distribution is
xeS
i=l
X, ■ -
n
(5.5.2)
and we recognize this last quantity to be the sample mean, x. The variance of the empirical distri-
bution is
cr2 = £(* - p)2fx(x) = - ^)2 ’ - (5.5.3)
xeS i— 1
and this last quantity looks very close to what we already know to be the sample variance.
= — '-r y (x, - x)2.
n — 1 i
i=\
(5.5.4)
The empirical quantile function is the inverse of the ECDF. See Section 6.3.1.
5.5.1 How to do it with R
The empirical distribution is not directly available as a distribution in the same way that the other
base probability distributions are, but there are plenty of resources available for the determined
investigator.
Given a data vector of observed values x, we can see the empirical CDF with the ecdf function:
> x <- c(4, 7, 9, 11, 12)
> ecdf(x)
Empirical CDF
Call: ecdf(x)
x [ 1 : 5] = 4, 7, 9, 11, 12
The above shows that the returned value of ecdf(x) is not a number but rather a function.
The ECDF is not usually used by itself in this form, by itself. More commonly it is used as an
intermediate step in a more complicated calculation, for instance, in hypothesis testing (see Chapter
10) or resampling (see Chapter 13). It is nevertheless instructive to see what the ecdf looks like,
and there is a special plot method for ecdf objects.
> plot (ecdf (x))
5.5. THE EMPIRICAL DISTRIBUTION
127
ecdf(x)
C=
LL
00
O
CD
O
■st
o
CVJ
o
o
o
X
Figure 5.5.1: The empirical CDF
See Figure 5.5.1. The graph is of a right-continuous function with jumps exactly at the locations
stored in x. There are no repeated values in x so all of the jumps are equal to 1/5 = 0.2.
The empirical PDF is not usually of particular interest in itself, but if we really wanted we could
define a function to serve as the empirical PDF:
> epdf <- function(x) function(t) {sum(x %in% t) /length(x) }
> x <- cfffl, 0, 1)
> epdf(x)CO) # should be 2/3
[1] 0.6666667
To simulate from the empirical distribution supported on the vector x, we use the sample
function.
> x <- c(6), 0, 1)
> sample(x, size = 7, replace = TRUE)
[1] 0 1 0 1 1 0 0
128
CHAPTER 5. DISCRETE DISTRIBUTIONS
We can get the empirical quantile function in R with quantile(x , probs = p, type = 1);
see Section 6.3.1.
As we hinted above, the empirical distribution is significant more because of how and where it
appears in more sophisticated applications. We will explore some of these in later chapters - see,
for instance. Chapter 13.
5.6 Other Discrete Distributions
The binomial and discrete uniform distributions are popular, and rightly so; they are simple and
form the foundation for many other more complicated distributions. But the particular uniform and
binomial models only apply to a limited range of problems. In this section we introduce situations
for which we need more than what the uniform and binomial offer.
5.6.1 Dependent Bernoulli Trials
The Hypergeometric Distribution
Consider an urn with 7 white balls and 5 black balls. Let our random experiment be to randomly
select 4 balls, without replacement, from the urn. Then the probability of observing 3 white balls
(and thus 1 black ball) would be
P(3W,1B) =
(5.6.1)
More generally, we sample without replacement K times from an urn with M white balls and N
black balls. Let X be the number of white balls in the sample. The PMF of X is
fx(x) =
(5.6.2)
We say that X has a hypergeometric distribution and write X ~ hyper(m = M, n = N, k = K).
The support set for the hypergeometric distribution is a little bit tricky. It is tempting to say that
x should go from 0 (no white balls in the sample) to K (no black balls in the sample), but that does
not work if K > M, because it is impossible to have more white balls in the sample than there were
white balls originally in the urn. We have the same trouble if K > N. The good news is that the
majority of examples we study have K < M and K < N and we will happily take the support to be
x = 0, 1, . .., K.
It is shown in Exercise 5.5 that
p = K
M
M + N ’
o-z = K-
MN M+N -K
'{M + N)1 M + N - 1 '
(5.6.3)
The associated R functions for the PMF and CDF are dhyperfx, m, n, k) and phyper,
respectively. There are two more functions: qhyper, which we will discuss in Section 6.3.1, and
rhyper, discussed below.
Example 5.20. Suppose in a certain shipment of 250 Pentium processors there are 17 defective
processors. A quality control consultant randomly collects 5 processors for inspection to determine
whether or not they are defective. Let X denote the number of defectives in the sample.
5. 6. OTHER DISCRETE DISTRIB UTIONS
129
1. Find the probability of exactly 3 defectives in the sample, that is, find P(W = 3).
Solution: We know that X ~ hyper(m = 17, n = 233, k = 5). So the required probability is
just
To calculate it in R we just type
> dhyper(3, m = 17, n = 233, k - 5)
[1] 0 . 002 3 5 1 1 5 B
To find it with the R Commander we go Probability > Discrete Distributions > Hyperge-
ometric distribution > Hypergeometric probabilities We fill in the parameters m = 17,
n = 233, and k — 5. Click OK, and the following table is shown in the window.
> A <- data, frame (Pr = dhyper(6:4, m = 17, n = 233, k = 5))
> rownames(A) <- 6:4
> A
0 7.011261e-Ql
1 2 . 602433e-01
2 3 . 620776e-02
3 2 . 351153e-03
4 7 . 093997e-05
We wanted P(A = 3), and this is found from the table to be approximately 0.0024. The value
is rounded to the fourth decimal place.
We know from our above discussion that the sample space should be x = 0, 1, 2, 3, 4, 5, yet,
in the table the probabilities are only displayed for x = 1,2,3, and 4. What is happening?
As it turns out, the R Commander will only display probabilities that are 0.00005 or greater.
Since x = 5 is not shown, it suggests that the outcome has a tiny probability. To find its exact
value we use the dhyper function:
> dhyper(5, m = 17, n = 233, k - 5)
[1] 7 . 916049e-07
In other words, P(X = 5) « 0.0000007916049, a small number indeed.
2. Find the probability that there are at most 2 defectives in the sample, that is, compute P(X <
2).
Solution: Since P(X < 2) = P(X = 0, 1,2), one way to do this would be to add the 0, 1,
and 2 entries in the above table, this gives 0.7011 + 0.2602 + 0.0362 = 0.9975. Our answer
should be correct up to the accuracy of 4 decimal places. However, a more precise method
Pr
130
CHAPTER 5. DISCRETE DISTRIBUTIONS
is provided by the R Commander. Under the Hypergeometric distribution menu we select
Hypergeometric tail probabilities. . . . We fill in the parameters m, n, and k as before, but
in the Variable value(s) dialog box we enter the value 2. We notice that the Lower tail
option is checked, and we leave that alone. Click OK.
> phyper(2, m = 17, n = 233, k - 5)
[1] 0.9975771
And thus P(A < 2) ~ 0.9975771. We have confirmed that the above answer was correct up
to four decimal places.
3. Find P(X > 1).
The table did not give us the explicit probability P(A = 5), so we can not use the table to
give us this probability. We need to use another method. Since P(A > 1) = 1 - P(X < 1) =
1 - Ex(l), we can find the probability with Hypergeometric tail probabilities We enter
1 for Variable Value(s), we enter the parameters as before, and in this case we choose the
Upper tail option. This results in the following output.
> phyper(l, m = 17, n = 233, k - 5, lower .tail = FALSE)
[1] 0.03863065
In general, the Upper tail option of a tail probabilities dialog computes P(A > x ) for all
given Variable Value(s) x.
4. Generate 100, 000 observations of the random variable X.
We can randomly simulate as many observations of X as we want in R Commander. Simply
choose Simulate hypergeometric variates. . . in the Hypergeometric distribution dialog.
In the Number of samples dialog, type 1. Enter the parameters as above. Under the Store
Values section, make sure New Data set is selected. Click OK.
A new dialog should open, with the default name Simsetl. We could change this if we like,
according to the rules for R object names. In the sample size box, enter 100000. Click OK.
In the Console Window, R Commander should issue an alert that Simsetl has been initial-
ized, and in a few seconds, it should also state that 100,000 hypergeometric variates were
stored in hyper, siml. We can view the sample by clicking the View Data Set button on
the R Commander interface.
We know from our formulas that p — K ■ M/(M + N) = 5 * 17/250 = 0.34. We can check
our formulas using the fact that with repeated observations of X we would expect about 0.34
defectives on the average. To see how our sample reflects the true mean, we can compute the
sample mean
Rcmdr> mean(Simset2$hyper . siml , na.rm=TRUE)
[1] 0.340344
Rcmdr> sd(Simset2$hyper . siml , na.rm=TRUE)
[1] 0.5584982
5. 6. OTHER DISCRETE DISTRIB UTIONS
131
We see that when given many independent observations of X, the sample mean is very close
to the true mean p. We can repeat the same idea and use the sample standard deviation to
estimate the true standard deviation of X. From the output above our estimate is 0.5584982,
and from our formulas we get
cr = K
MN M+N-K
(M + N)2 M + N - 1
0.3117896,
with cr = yfcr2 » 0.558381 1944. Our estimate was pretty close.
From the console we can generate random hypergeometric variates with the rhyper function,
as demonstrated below.
> rhyper(16 , m = 17, n = 233, k = 5)
[1] 00800208®!
Sampling With and Without Replacement
Suppose that we have a large urn with, say, M white balls and N black balls. We take a sample of
size n from the urn, and let X count the number of white balls in the sample. If we sample
without replacement, then X ~ hyper(m -M, n = N, k = n) and has mean and variance
M
n ,
M + N
MN M + N -n
U (M + N)2 M + N - 1 ’
M / M \M + N — n
U M + N l1 ” M + N) M + N-V
On the other hand, if we sample
with replacement, then X ~ binom(size = n, prob = M/(M + N)) with mean and variance
M
n ,
M + N
M( M \
n 1 .
M + N\ M + N)
We see that both sampling procedures have the same mean, and the method with the larger variance
is the “with replacement” scheme. The factor by which the variances differ.
M + N - n
M + N-V
(5.6.4)
is called a. finite population correction. For a fixed sample size n, as M,N — > oo it is clear that the
correction goes to 1, that is, for infinite populations the sampling schemes are essentially the same
with respect to mean and variance.
132
CHAPTER 5. DISCRETE DISTRIBUTIONS
5.6.2 Waiting Time Distributions
Another important class of problems is associated with the amount of time it takes for a specified
event of interest to occur. For example, we could flip a coin repeatedly until we observe Heads. We
could toss a piece of paper repeatedly until we make it in the trash can.
The Geometric Distribution
Suppose that we conduct Bernoulli trials repeatedly, noting the successes and failures. Let X be the
number of failures before a success. If PCS') = p then X has PMF
fx(x) = p(]~ p)x, x = 0,1,2,... (5.6.5)
(Why?) We say that X has a Geometric distribution and we write X ~ geom(prob = p). The
associated R functions are dgeom(x, prob), pgeom, qgeom, and rhyper, which give the PMF,
CDF, quantile function, and simulate random variates, respectively.
Again it is clear that f(x) > 0 and we check that £ fix) = 1 (see Equation E.3.9 in Appendix
E.3):
YjPV-pY
x=0
~-pJ]qx
x=0
= 1.
We will find in the next section that the mean and variance are
P =
1 ~P
P
— and cr2
P
(5.6.6)
Example 5.21. The Pittsburgh Steelers place kicker, Jeff Reed, made 81.2% of his attempted field
goals in his career up to 2006. Assuming that his successive field goal attempts are approximately
Bernoulli trials, find the probability that Jeff misses at least 5 field goals before his first successful
goal.
Solution : If X - the number of missed goals until Jeff’s first success, then X ~ geomfprob =
0.812) and we want P(A > 5) = IP (A > 4). We can find this in R with
> pgeom(4, prob = SI. 812, lower, tail = FALSE )
[1] 8.8882348493
Note 5.22. Some books use a slightly different definition of the geometric distribution. They con-
sider Bernoulli trials and let Y count instead the number of trials until a success, so that Y has
PMF
friy) = p( 1 ~PY~\ y= 1,2,3,... (5.6.7)
When they say “geometric distribution”, this is what they mean. It is not hard to see that the
two definitions are related. In fact, if X denotes our geometric and Y theirs, then Y = X + 1.
Consequently, they have pY - Px + 1 and cr2 = cr2x.
5. 6. OTHER DISCRETE DISTRIB UTIONS
133
The Negative Binomial Distribution
We may generalize the problem and consider the case where we wait for more than one success.
Suppose that we conduct Bernoulli trials repeatedly, noting the respective successes and failures.
Let X count the number of failures before r successes. If P(S) = p then X has PMF
fx(x) = * l ^/(l ~P)X, x = 0,1,2,. . . (5.6.8)
We say that X has a Negative Binomial distribution and write X ~ nbinom(size = r, prob = p).
The associated R functions are dnbinora(x, size, prob), pnbinom, qnbinom, and rnbinom,
which give the PMF, CDF, quantile function, and simulate random variates, respectively.
As usual it should be clear that fx(x') > 0 and the fact that £ fx(x) = 1 follows from a general-
ization of the geometric series by means of a Maclaurin’s series expansion:
1
1^7
1
(1 - tY
for - 1 < t < 1 , and
for -1 < t < 1.
(5.6.9)
(5.6.10)
Therefore
Yjfx{x) = p'Yj
r + x — 1
r- 1
q* = pr(l - q)~r = 1,
since \q\ = [1 - p\ < 1.
(5.6.11)
Example 5.23. We flip a coin repeatedly and let X count the number of Tails until we get seven
Heads. What is F(X = 5)?
Solution: We know that A ~ nbinom(size = 7, prob = 1/2).
P(A = 5) = fx( 5) = + 5_ : 1 j(l/2)7(l/2)5 = J161J2-12
and we can get this in R with
> dnbinom(5 , size = 7, prob = (9.5)
[1] 0.112793®
Let us next compute the MGF of X ~ nbinom(size = r, prob = p).
-p'( 1 - qe ') r , provided \qe'\ < 1,
134
CHAPTER 5. DISCRETE DISTRIBUTIONS
and so
Mx(t) = (w?) ’ f°rqS' < L (5-6.12)
We see that qe' < 1 when t < - ln(l - p).
Let X ~ nbinom(size = r, prob = p) with M(t) = pr( 1 - qe')~r. We proclaimed above the
values of the mean and variance. Now we are equipped with the tools to find these directly.
M'{t)=p\-r){\-qQtr-\-q^),
-rqe p ( 1 - qe)
rqe1
1 - qe
, rq rq
M (0) =^-L- • 1 = — .
t\—r— 1
- M(t ), and so
1 -q
Thus p = rqlp. We next find EX2.
m-( o) =
(1 - ge*)
= rqp + ''g2 J + ^
rqe
1 - ge
l/=0
P \P
= rq I ,-q\2
P2 \pj
Finally we may say cr2 = M"( 0) - [M'(0)]2 = rqlp2.
Example 5.24. A random variable has MGF
. 0.19
Mx(t) =
ThenX ~ nbinom(size = 31, prob = 0.19).
1 - 0.81e'
31
Note 5.25. As with the Geometric distribution, some books use a slightly different definition of the
Negative Binomial distribution. They consider Bernoulli trials and let Y be the number of trials
until r successes, so that Y has PMF
fr(y) = _ |W(1 - P)y~\ y - r, r + 1, r + 2, . . . (5.6.13)
It is again not hard to see that if X denotes our Negative Binomial and Y theirs, then Y — X + r.
Consequently, they have py - Px + r and cr2 = cr2,.
5.6.3 Arrival Processes
The Poisson Distribution
This is a distribution associated with “rare events”, for reasons which will become clear in a mo-
ment. The events might be:
5. 6. OTHER DISCRETE DISTRIB UTIONS
135
• traffic accidents,
• typing errors, or
• customers arriving in a bank.
Let A be the average number of events in the time interval [0,1], Let the random variable X count
the number of events occurring in the interval. Then under certain reasonable conditions it can be
shown that
We use the notation X ~ pois(lambda = d). The associated R functions are dpoisfx, lambda),
ppois, qpois, and rpois, which give the PMF, CDF, quantile function, and simulate random
variates, respectively.
What are the reasonable conditions? Divide [0, 1] into subintervals of length l/n. A Poisson
process satisfies the following conditions:
• the probability of an event occurring in a particular subinterval is » A/ii.
• the probability of two or more events occurring in any subinterval is ~ 0.
• occurrences in disjoint subintervals are independent.
Remark 5.26. If X counts the number of events in the interval [0,/] and A is the average number
that occur in unit time, then X ~ poisllambda = At), that is,
Example 5.27. On the average, five cars arrive at a particular car wash every hour. Let X count the
number of cars that arrive from 10AM to 1 1AM. Then X ~ poisllambda = 5). Also, /; = cr2 = 5.
What is the probability that no car arrives during this period?
Solution: The probability that no car arrives is
Example 5.28. Suppose the car wash above is in operation from 8AM to 6PM, and we let Y be the
number of customers that appear in this period. Since this period covers a total of 10 hours, from
Remark 5.26 we get that Y ~ pois(lambda = 5*10 = 50). What is the probability that there are
between 48 and 50 customers, inclusive?
Solution: We want P(48 < Y < 50) = P(X < 50) - P(X < 47).
(5.6.14)
(5.6.15)
5°
P(A = 0) = e~5 —
= e~5 * 0.0067.
> diff (ppois (c(47 , 56)), lambda = 56)))
[1] 0.1678485
136
CHAPTER 5. DISCRETE DISTRIBUTIONS
5.7 Functions of Discrete Random Variables
We have built a large catalogue of discrete distributions, but the tools of this section will give us the
ability to consider infinitely many more. Given a random variable X and a given function h, we may
consider Y — h(X). Since the values of X are determined by chance, so are the values of Y. The
question is, what is the PMF of the random variable T? The answer, of course, depends on h. In
the case that h is one-to-one (see Appendix E.2), the solution can be found by simple substitution.
Example 5.29. Let X ~ nbinom(size = r, prob = p). We saw in 5.6 that X represents the
number of failures until r successes in a sequence of Bernoulli trials. Suppose now that instead we
were interested in counting the number of trials (successes and failures) until the rth success occurs,
which we will denote by Y. In a given performance of the experiment, the number of failures (X)
and the number of successes (r) together will comprise the total number of trials (T), or in other
words, X + r - Y. We may let h be debited by h(x) — x + r so that Y = h(X), and we notice that
h is linear and hence one-to-one. Finally, X takes values 0, 1, 2, . . . implying that the support of Y
would be {r, r + 1, r + 2, . . .}. Solving for X we get X = Y - r. Examining the PMF of X
fx(x) = | r+J.X_l - P)X’ (5.7.1)
we can substitute x = y - r to get
fr(y ) = fx(y - r),
(r + (y - r) - 1
r- 1
>'- 1
r- 1
p\ i - pT\
pr( 1 - p)y~r, y - r, r + 1 , . . .
Even when the function h is not one-to-one, we may still hnd the PMF of Y simply by accumu-
lating, for each y, the probability of all the jc’s that are mapped to that y.
Proposition 5.30. Let X be a discrete random variable with PMF fx supported on the set S x- Let
Y — h(X) for some function h. Then Y has PMF fy defined by
fr(y ) = Yj fx(x)
{jreSxl h(x)=y }
(5.7.2)
Example 5.31. Let X ~ binom(size = 4, prob = 1/2), and let Y = (X - l)2. Consider the
following table:
X
0
1
2
3
4
fx(x)
1/16
1/4
6/16
1/4
1/16
y = (x-2)2
1
0
1
4
9
From this we see that Y has support S y - {0, 1,4,9). We also see that h(x) = (x - l)2 is not one-
to-one on the support of X, because both x = 0 and x — 2 are mapped by h to y - 1. Nevertheless,
we see that Y — 0 only when X = 1, which has probability 1/4; therefore, fy( 0) should equal 1/4.
5.7. FUNCTIONS OF DISCRETE RANDOM VARIABLES
137
A similar approach works for y — 4 and y — 9. And Y — 1 exactly when X — 0 or X — 2, which has
total probability 7/16. In summary, the PMF of Y may be written:
y 0 14 9
fx(x) 1/4 7/16 1/4 1/16
Note that there is not a special name for the distribution of Y, it is just an example of what to do
when the transformation of a random variable is not one-to-one. The method is the same for more
complicated problems.
Proposition 5.32. IfX is a random variable with E X — /i and Var(X) = cr2, then the mean and
variance ofY — mX + b is
[iY = mu + b, o-y = m2cr2, crY - \m\cr. (5.7.3)
138
CHAPTER 5. DISCRETE DISTRIBUTIONS
Chapter Exercises
Exercise 5.1. A recent national study showed that approximately 44.7% of college students have
used Wikipedia as a source in at least one of their term papers. Let X equal the number of students
in a random sample of size n — 31 who have used Wikipedia as a source.
1 . How is X distributed?
X ~ binom(size = 31, prob = 0.447)
2. Sketch the probability mass function (roughly).
3. Sketch the cumulative distribution function (roughly).
5. 7. FUNCTIONS OF DISCRETE RANDOM VARIABLES
139
Binomial Dist'n: Trials = 31, Prob of success = 0.447
_Q
OS
.Q
O
:
Q_
<D
>
H — 1
jd
E
O
5 10 15 20 25
Number of Successes
4. Find the probability that X is equal to 17.
> dbinom(17 , size = 31, prob = 6.447)
[1] 0 . 07532248
5. Find the probability that X is at most 13.
> pbinom(13 , size = 31, prob = 6.447)
[1] 0.451357
6. Find the probability that X is bigger than 11.
> pbinom(ll , size = 31, prob = 6.447, lower .tail = FALSE)
[1] 0.8020339
7. Find the probability that X is at least 15.
> pbinom(14 , size = 31, prob = 6.447, lower . tail = FALSE)
[1] 0.406024
8. Find the probability that X is between 16 and 19, inclusive.
140
CHAPTER 5. DISCRETE DISTRIBUTIONS
> sum(dbinom(16: 19, size = 31, prob = IS. 447))
[1] 8.2544758
> di£f(pbinom(c(19, 15), size = 31, prob = 6.447, lower . tail = FALSE))
[1] 8.2544758
9. Give the mean of X, denoted E X.
> library(distrEx)
> X = Binom(size = 31, prob = IS. 447)
> E(X)
[1] 13.857
10. Give the variance of X.
> var(X)
[1] 7.662921
1 1 . Give the standard deviation of X.
> sd(X)
[1] 2.768198
12. Find E(4X + 51.324)
> E(4 * X + 51.324)
[1] 186.752
Exercise 5.2. For the following situations, decide what the distribution of X should be. In nearly
every case, there are additional assumptions that should be made for the distribution to apply;
identify those assumptions (which may or may not hold in practice.)
1. We shoot basketballs at a basketball hoop, and count the number of shots until we make a
goal. Let X denote the number of missed shots. On a normal day we would typically make
about 37% of the shots.
2. In a local lottery in which a three digit number is selected randomly, let X be the number
selected.
3. We drop a Styrofoam cup to the floor twenty times, each time recording whether the cup
comes to rest perfectly right side up, or not. Let X be the number of times the cup lands
perfectly right side up.
5.7. FUNCTIONS OF DISCRETE RANDOM VARIABLES
141
4. We toss a piece of trash at the garbage can from across the room. If we miss the trash can,
we retrieve the trash and try again, continuing to toss until we make the shot. Let X denote
the number of missed shots.
5. Working for the border patrol, we inspect shipping cargo as when it enters the harbor looking
for contraband. A certain ship comes to port with 557 cargo containers. Standard practice
is to select 10 containers randomly and inspect each one very carefully, classifying it as
either having contraband or not. Let X count the number of containers that illegally contain
contraband.
6. At the same time every year, some migratory birds land in a bush outside for a short rest. On
a certain day, we look outside and let X denote the number of birds in the bush.
7. We count the number of rain drops that fall in a circular area on a sidewalk during a ten
minute period of a thunder storm.
8. We count the number of moth eggs on our window screen.
9. We count the number of blades of grass in a one square foot patch of land.
10. We count the number of pats on a baby’s back until (s)he burps.
Exercise 5.3. Show that E(A - p)2 = E X2 -/r. Hint: expand the quantity (X - p)2 and distribute
the expectation over the resulting terms.
Exercise 5.4. If X ~ binom(size = n, prob = p) show that IEX(X - 1) = n(n - 1 )p2.
Exercise 5.5. Calculate the mean and variance of the hypergeometric distribution. Show that
M , MN M+N-K
h M + N (M + N)2M + N- 1
(5.7.4)
142
CHAPTER 5. DISCRETE DISTRIBUTIONS
Chapter 6
Continuous Distributions
The focus of the last chapter was on random variables whose support can be written down in a list
of values ( finite or countably infinite), such as the number of successes in a sequence of Bernoulli
trials. Now we move to random variables whose support is a whole range of values, say, an interval
( a,b ). It is shown in later classes that it is impossible to write all of the numbers down in a list;
there are simply too many of them.
This chapter begins with continuous random variables and the associated PDFs and CDFs The
continuous uniform distribution is highlighted, along with the Gaussian, or normal, distribution.
Some mathematical details pave the way for a catalogue of models.
The interested reader who would like to learn more about any of the assorted discrete distribu-
tions mentioned below should take a look at Continuous Univariate Distributions, Volumes 1 and
2 by Johnson et al [47, 48].
What do I want them to know?
• how to choose a reasonable continuous model under a variety of physical circumstances
• basic correspondence between continuous versus discrete random variables
• the general tools of the trade for manipulation of continuous random variables, integration,
etc.
• some details on a couple of continuous models, and exposure to a bunch of other ones
• how to make new continuous random variables from old ones
6.1 Continuous Random Variables
6.1.1 Probability Density F unctions
Continuous random variables have supports that look like
Sx = [a,b] or (a, b), (6.1.1)
143
144
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
or unions of intervals of the above form. Examples of random variables that are often taken to be
continuous are:
• the height or weight of an individual,
• other physical measurements such as the length or size of an object, and
• durations of time (usually).
Every continuous random variable X has a probability density function (PDF) denoted fx associated
with it1 that satisfies three basic properties:
1- fx(x) > 0 for x E Sx,
2. JxeS x fx(X) d' = 1 ' alUl
3. P(X e A) — fxeA fx(x) djr, for an event A c S x-
Remark 6.1. We can say the following about continuous random variables:
• Usually, the set A in 3 takes the form of an interval, for example, A = [c, d], in which case
P(X e A) = J fx(x) dx. (6.1.2)
• It follows that the probability that X falls in a given interval is simply the area under the
curve of fx over the interval.
• Since the area of a line x = c in the plane is zero, IP (A" = c) - 0 for any value c. In other
words, the chance that X equals a particular value c is zero, and this is true for any number
c. Moreover, when a < b all of the following probabilities are the same:
P(a < X < b) = P(a < X < b) = P(a < X < b) = P(a < X < b). (6.1.3)
• The PDF fx can sometimes be greater than 1 . This is in contrast to the discrete case; every
nonzero value of a PMF is a probability which is restricted to lie in the interval [0,1],
We met the cumulative distribution function, Fx , in Chapter 5. Recall that it is defined by Fx(f) =
P(X < f), for -oo < t < oo. While in the discrete case the CDF is unwieldy, in the continuous case
the CDF has a relatively convenient form:
Fx(t) = PCX <t) =
fx(x) dx.
—oo < t < oo.
(6.1.4)
Remark 6.2. For any continuous CDF Fx the following are true.
• Fx is nondecreasing , that is, t\ < implies Fx(t\) < Fxfo)-
Not true. There are pathological random variables with no density function. (This is one of the crazy things that
can happen in the world of measure theory). But in this book we will not get even close to these anomalous beasts, and
regardless it can be proved that the CDF always exists.
6.1. CONTINUOUS RANDOM VARIABLES
145
• Fx is continuous (see Appendix E.2). Note the distinction from the discrete case: CDFs of
discrete random variables are not continuous, they are only right continuous.
• lim^-oo Fx(t) = 0 and limf_oo Fx(t) = 1 .
There is a handy relationship between the CDF and PDF in the continuous case. Consider the
derivative of Fx:
F'x(t) = ^-Fx(t) = £ f fx(x)dx = fx(t), (6.1.5)
dr dr
the last equality being true by the Fundamental Theorem of Calculus, part (2) (see Appendix E.2).
In short, (Fx)' = fx in the continuous case2.
6.1.2 Expectation of Continuous Random Variables
For a continuous random variable X the expected value of g(X) is
f
x&S
Egd 0
g(x)fx(x) dx,
(6.1.6)
provided the (potentially improper) integral fs |g(x)| f(x)dx is convergent. One important example
is the mean /i, also known as IE X:
H = El
-f
J xeS
xfx(x) dx,
provided fs \x\f(x)dx is finite. Also there is the
variance
o" = E(X - /r)‘
(x - fi)-fx(x) dx,
(6.1.7)
(6.1.8)
which can be computed with the alternate formula cr2 = EX2 - (EX)2. In addition, there is the
standard deviation cr = Vcr2. The moment generating function is given by
Mx(t) = E e
tx
etxfx(x) dx,
(6.1.9)
provided the integral exists (is finite) for all t in a neighborhood of t = 0.
Example 6.3. Let the continuous random variable X have PDF
fx(x) = 3x2, 0 < x < 1.
poo
We will see later that fx belongs to the Beta family of distributions. It is easy to see that J /(x)dx =
1.
Xoo /-*\
fx(x)dx =
oo J 0
3x2 dx
= x
i
x=0
= l3 - 03
= 1.
“In the discrete case, fx(x ) = Fx(x ) - Fx(t).
146
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
This being said, we may find P(0.14 < X < 0.71).
✓>0.71
P(0.14 < X < 0.71) = 3x1dx,
Jo. 14
3 10-7 1
•*' Lv=0. 14
= 0.713 - 0.143
« 0.355167.
We can find the mean and variance in an identical manner.
P
XOO 1
xfx(x)dx - x ■ 3xz dx,
OO 0
3
4'X
_ 3
~ 4'
.4,1
l.v=0’
It would perhaps be best to calculate the variance with the shortcut formula crz - IE X2 - \r\
XOO A»1
x2fx(x)dx = I x2 ■ 3x2 dv
OO xJ 0
3 I1
J .5
= 5X
= 3/5.
x=0
which gives cr2 = 3/5 - (3/4)2 = 3/80.
Example 6.4. We will try one with unbounded support to brush up on improper integration. Let
the random variable X have PDF
x>i-
X4 9
XOO
f(x)dx = 1 :
r fx(x)dx = r
J- oo J 1
= lim r
1
— dr
x4
x4
d.Y
= lim 3 — x
t—>oo —3
= - lim -7-1
= 1.
X=1
6.1. CONTINUOUS RANDOM VARIABLES
147
We calculate P(3.4 < X < 7.1):
P(3.4 <X<1
,-r
3x 4dx
1 74
= 3
-3 x=3.4
= -1(7. r3 -3.4~3)
~ 0.0226487123.
We locate the mean and variance just like before.
/OO /"»CX
xfx(x)dx = I
CO J 1
x • — dx
x4
= - 1 f lim i - 1
2 yr— >°° (2
_ 3
“ 2'
Again we use the shortcut <x2 = E X2 - yir :
Xoo /-*&.
x2fx(x)dx =
CO Jl
x2 • — dx
X4
- 3
= -3 ( lim — — 1
y->oo
= 3,
which closes the example with cr~ - 3 - (3/2)“ = 3/4.
6.1.3 How to do it with R
There exist utilities to calculate probabilities and expectations for general continuous random vari-
ables, but it is better to find a built-in model, if possible. Sometimes it is not possible. We show
how to do it the long way, and the distr package way.
Example 6.5. Let X have PDF fix) = 3x2, 0 < x < 1 and find P(0.14 < X < 0.71). (We will
ignore that X is a beta random variable for the sake of argument.)
> f <- function(x) 3 * xA2
> integrate(f, lower = (S.14, upper = 6.71)
0.355167 with absolute error < 3.9e-15
Compare this to the answer we found in Example 6.3. We could integrate the function x/(x) =
3*xA3 from zero to one to get the mean, and use the shortcut <x2 =EI2- (EX)2 for the variance.
148
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
Example 6.6. Let X have PDF f(x) = 3/x4, x > 1. We may integrate the function xf(x) — 3/xA3
from zero to infinity to get the mean of X.
> g <- function(x) 3/xA3
> integrate (g , lower = 1, upper = Inf)
1.5 with absolute error < 1.7e-14
Compare this to the answer we got in Example 6.4. Use -Inf for -oo.
Example 6.7. Let us redo Example 6.3 with the distr package. The method is similar to that
encountered in Section 5.1.3 in Chapter 5. We define an absolutely continuous random variable:
> library (distr)
> f <- function(x) 3 * xA2
> X <- AbscontDistribution(d = f, lowl = 0, upl = 1)
> p(X) (S. 71) - p(X) (0. 14)
[1] 0.355167
Compare this answers we found earlier. Now let us try expectation with the distrEx package
[74]:
> library (distrEx)
> E(X)
[1] 0.7496337
> var(X)
[1] 0.03768305
> 3/80
[1] 0.0375
Compare these answers to the ones we found in Example 6.3. Why are they different? Because
the distrEx package resorts to numerical methods when it encounters a model it does not recog-
nize. This means that the answers we get for calculations may not exactly match the theoretical
values. Be careful.
6.2 The Continuous Uniform Distribution
A random variable X with the continuous uniform distribution on the interval (a, b) has PDF
1
fx(x) = , a < x < b.
b - a
(6.2.1)
The associated R function is dunif(min = a, max = b ). We write X ~ unif(min = a , max = b).
Due to the particularly simple form of this PDF we can also write down explicitly a formula for the
CDF Fx:
0,
t-a
b-a ’
l,
Fx(t) =
t < 0,
a < t < b,
t > b.
(6.2.2)
6.3. THE NORMAL DISTRIBUTION
149
The continuous uniform distribution is the continuous analogue of the discrete uniform distribution;
it is used to model experiments whose outcome is an interval of numbers that are “equally likely” in
the sense that any two intervals of equal length in the support have the same probability associated
with them.
Example 6.8. Choose a number in [0,1] at random, and let X be the number chosen. Then X ~
unif(rain = 0, max = 1).
The mean of X ~ unif(min = a. max = b) is relatively simple to calculate:
yu = E X = I x fx(x) dx,
I
■r
i
b — a
dx.
1 x2
b — a 2
1 b2 - a
x—a
2 „2
b — a 2
b + a
2 ’
using the popular formula for the difference of squares. The variance is left to Exercise 6.4.
6.3 The Normal Distribution
We say that X has a normal distribution if it has PDF
fx(x) = 1 exp | (X2(t2^ | , -oo < x < oo. (6.3.1)
We write X ~ norm(mean - /a. sd = cr), and the associated R function is dnorm(x, mean = ®,
sd = 1).
The familiar bell-shaped curve, the normal distribution is also known as the Gaussian dis-
tribution because the German mathematician C. F. Gauss largely contributed to its mathematical
development. This distribution is by far the most important distribution, continuous or discrete.
The normal model appears in the theory of all sorts of natural phenomena, from to the way parti-
cles of smoke dissipate in a closed room, to the journey of a bottle in the ocean to the white noise
of cosmic background radiation.
When /r = 0 and cr = 1 we say that the random variable has a standard normal distribution and
we typically write Z ~ norm(mean = 0, sd = 1). The lowercase Greek letter phi Up) is used to
denote the standard normal PDF and the capital Greek letter phi <t> is used to denote the standard
normal CDF: for -oo < z < oo,
1 nt
(p(z) = —= e~;2/2 and d>(f) = I cp(z) dz.
y2n J-o o
(6.3.2)
150
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
Proposition 6.9. IfX ~ norm(mean - /j, sd- cr) then
Z = — — - ~ norm(mean = 0, sd = 1). (6.3.3)
cr
The MGF of Z ~ norm(mean = 0, sd = 1) is relatively easy to derive:
Mz(t)
dz,
= e^f f J-e-IH^I,
\J-00 V27T /
and the quantity in the parentheses is the total area under a norm(mean = —t, sd = 1) density,
which is one. Therefore,
Mz{t) = eTr^2, -oo < t < co. (6.3.4)
Example 6.10. The MGF of X ~ norm(mean = //, sd = cr) is then not difficult either because
X-u
Z = , or rewriting, X = crZ + p.
cr
Therefore:
Mx(t) = IEe'x = Ee1^ = EertV = e'*' Mz(crt ),
and we know that Mz(t ) = e'"^2, thus substituting we get
Mx(t) = e,/Je(crr)2/2 = exp {pr + cr2f2/2} ,
for -co < t < co.
Fact 6.11. Ihe same argument above shows that ifX has MGF Mx(t ) then the MGF ofY — a + bX
is
MY(t) = e,aMx(bt). (6.3.5)
Example 6.12. The 68-95-99.7 Rule. We saw in Section 3.3.6 that when an empirical distribution
is approximately bell shaped there are specific proportions of the observations which fall at varying
distances from the (sample) mean. We can see where these come from - and obtain more precise
proportions - with the following:
> pnorm(l : 3) - pnorm(- (1 : 3))
[1] 0.6826895 0.9544997 0.9973002
Example 6.13. Let the random experiment consist of a person taking an IQ test, and let X be the
score on the test. The scores on such a test are typically standardized to have a mean of 100 and a
standard deviation of 15. What is P(85 < X < 1 15)?
Solution: this one is easy because the limits 85 and 115 fall exactly one standard deviation
(below and above, respectively) from the mean of 100. The answer is therefore approximately
68%.
6.3. THE NORMAL DISTRIBUTION
151
6.3.1 Normal Quantiles and the Quantile Function
Until now we have been given two values and our task has been to find the area under the PDF
between those values. In this section, we go in reverse: we are given an area, and we would like to
find the value(s) that correspond to that area.
Example 6.14. Assuming the IQ model of Example 6.13, what is the lowest possible IQ score that
a person can have and still be in the top 1% of all IQ scores?
Solution: If a person is in the top 1%, then that means that 99% of the people have lower IQ
scores. So, in other words, we are looking for a value x such that Fix) = \P(X < x) satisfies F(x) =
0.99, or yet another way to say it is that we would like to solve the equation F(x)~ 0.99 = 0. For the
sake of argument, let us see how to do this the long way. We define the function g(x) = F(x) - 0.99,
and then look for the root of g with the uni root function. It uses numerical procedures to find
the root so we need to give it an interval of x values in which to search for the root. We can get
an educated guess from the Empirical Rule 3.13; the root should be somewhere between two and
three standard deviations (15 each) above the mean (which is 100).
> g <- function(x) pnorm(x, mean = ISIS), sd = 15) - S.99
> uniroot (g , interval = c(13&, 145))
$root
[1] 134.8952
$f .root
[1] -4 . 873®83e-09
Siter
[1] 6
$estim.prec
[1] 6.1®3516e-®5
The answer is shown in $root which is approximately 134.8952, that is, a person with this IQ
score or higher falls in the top 1% of all IQ scores.
The discussion in example 6.14 was centered on the search for a value x that solved an equation
F(x) - p, for some given probability p, or in mathematical parlance, the search for F 1 , the inverse
of the CDF of X, evaluated at p. This is so important that it merits a definition all its own.
Definition 6.15. The quantile function 3 of a random variable X is the inverse of its cumulative
distribution function:
Qx(p) = min {x : Fx(x) > p\ , 0 < p < 1. (6.3.6)
Remark 6.16. Here are some properties of quantile functions:
1 . The quantile function is defined and finite for all 0 < p < 1 .
"the precise definition of the quantile function is Qx(p) = inf \x '■ Fx(X) ^ p I, so at least it is well defined (though
perhaps infinite) for the values p = 0 and p = 1 .
152
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
2. Qx is left-continuous (see Appendix E.2). For discrete random variables it is a step function,
and for continuous random variables it is a continuous function.
3. In the continuous case the graph of Qx may be obtained by reflecting the graph of Fx about
the line y = x. In the discrete case, before reflecting one should: 1) connect the dots to get
rid of the jumps - this will make the graph look like a set of stairs, 2) erase the horizontal
lines so that only vertical lines remain, and finally 3) swap the open circles with the solid
dots. Please see Figure 5.3.2 for a comparison.
4. The two limits
lim Qx(p) and lim Qx(p)
p—>0+ p—*^~
always exist, but may be infinite (that is, sometimes lim^o Q(p) = ~°° and/or i Q( p) =
oo).
As the reader might expect, the standard normal distribution is a very special case and has its
own special notation.
Definition 6.17. For 0 < a < 1, the symbol za denotes the unique solution of the equation P(Z >
Za ) = o', where Z ~ norm(mean = 0, sd = 1). It can be calculated in one of two equivalent ways:
qnorm(l - a) and qnorm(ar, lower. tail = FALSE).
There are a few other very important special cases which we will encounter in later chapters.
6.3.2 How to do it with R
Quantile functions are defined for all of the base distributions with the q prefix to the distribution
name, except for the ECDF whose quantile function is exactly the Qx(p) =quantile(x, probs
= p , type = 1) function.
Example 6.18. Back to Example 6.14, we are looking for Qx( 0.99), where X ~ norm(mean =
100, sd = 15). It could not be easier to do with R.
> qnorm(6).99, mean = 16)61, sd = 15)
[1] 134.8952
Compare this answer to the one obtained earlier with uniroot.
Example 6.19. Find the values Z0.025, to. 01 . and zo.005 (these will play an important role from Chap-
ter 9 onward).
> qnorm(c (0.6)25, 6). 6)1, 61.6)6)5), lower .tail = FALSE)
[1] 1.959964 2.326348 2.575829
Note the lower . tail argument. We would get the same answer with
qnorm(c (0 . 975 , 0.99, 0.995))
6.4. FUNCTIONS OF CONTINUOUS RANDOM VARIABLES
153
6.4 Functions of Continuous Random Variables
The goal of this section is to determine the distribution of U — g(X) based on the distribution of
X. In the discrete case all we needed to do was back substitute for x = g '(«) in the PMF of X
(sometimes accumulating probability mass along the way). In the continuous case, however, we
need more sophisticated tools. Now would be a good time to review Appendix E.2.
6.4.1 The PDF Method
Proposition 6.20. Let X have PDF fx and let g be a function which is one-to-one with a differen-
tiable inverse g_1. Then the PDF of U — g(X) is given by
fu(u) = fx [g \u)]
(6.4.1)
Remark 6.21. The formula in Equation 6.4.1 is nice, but does not really make any sense. It is better
to write in the intuitive form
fu(u) = fx(x)
dx
d u
(6.4.2)
Example 6.22. Let X ~ norm(mean = /i, sd = cr), and let Y = ex. What is the PDF of T?
Solution: Notice first that eA > 0 for any x, so the support of Y is (0, oo). Since the transforma-
tion is monotone, we can solve y = eA for x to get x — In y, giving dx/dy = 1 /y. Therefore, for any
y > 0,
fr(y) = fxQny) ■
1
y
■ sfht
exp
(In y-p)2}
2cr2 J
1
y'
where we have dropped the absolute value bars since y > 0. The random variable Y is said to have
a lognormal distribution ; see Section 6.5.
Example 6.23. Suppose X ~ norm(mean = 0, sd = 1) and let Y — 4 - 3X. What is the PDF of Y?
The support of X is (-oo, oo), and as x goes from -oo to oo, the quantity y = 4 - 3x also traverses
(— oo, oo). Solving for x in the equation y = 4 - 3x yields x = — (y - 4)/3 giving dx/dy = -1/3. And
since
j ,
fx(x) = — — e~x~/2, -oo < x < oo,
\2n
we have
fr(y)
1 c-(y-4)2/2-32
3 V2 n
— oo < y < oo,
— oo < y < oo.
We recognize the PDF of Y to be that of a norm(raean = 4, sd = 3) distribution. Indeed, we may
use an identical argument as the above to prove the following fact:
Fact 6.24. If X ~ norm(mean = n, sd = cr) and ifY = a + bXfor constants a and b, with b 4 0,
then Y ~ norm(mean = a + bfi, sd = \b\cr).
154
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
Note that it is sometimes easier to postpone solving for the inverse transformation x = x{u).
Instead, leave the transformation in the form u = u(x) and calculate the derivative of the original
transformation
Au/Ax = g'(x). (6.4.3)
Once this is known, we can get the PDF of U with
fu(u) = fx(x)
1
(6.4.4)
An /Ax
In many cases there are cancellations and the work is shorter. Of course, it is not always true that
dx 1
Au Au /Ax’
(6.4.5)
but for the well-behaved examples in this book the trick works just fine.
Remark 6.25. In the case that g is not monotone we cannot apply Proposition 6.20 directly. How-
ever, hope is not lost. Rather, we break the support of X into pieces such that g is monotone on
each one. We apply Proposition 6.20 on each piece, and finish up by adding the results together.
6.4.2 The CDF method
We know from Section 6.1 that fx = F'x in the continuous case. Starting from the equation FY(y) =
P(T < vj, we may substitute g(X) for Y, then solve for X to obtain IP [A" < g 1 ( v)] , which is just
another way to write Fx\g _1(y)]. Differentiating this last quantity with respect to y will yield the
PDF of Y.
Example 6.26. Suppose X ~ uniffmin = 0, max = 1) and suppose that we let Y = - In X. What is
the PDF of F?
The support set of X is (0, 1), and y traverses (0, oo) as x ranges from 0 to 1, so the support set
of Y is S y = (0, oo). For any y > 0, we consider
FY{y) = P(F < >■) = P(- In X <y) - P(X > e^v) = 1 - P(X < e^-v),
where the next to last equality follows because the exponential function is monotone (this point will
be revisited later). Now since X is continuous the two probabilities Pi X < e -y) and P(X < e_v) are
equal; thus
1 - P(X < e-y) = 1 - P(X < e~y) = 1 - Fx(e~y).
Now recalling that the CDF of a unif(min = 0, max =1) random variable satisfies F(u) = u (see
Equation 6.2.2), we can say
FY(y) = 1 - Fx(eTy) = 1 - e"\ for y > 0.
We have consequently found the formula for the CDF of T; to obtain the PDF fY we need only
differentiate FY :
fY(y) ~ T” (1 ~ e v) = 0 - e--v(-l),
dv
°r fy(y ) = e v for y > 0. This turns out to be a member of the exponential family of distributions,
see Section 6.5.
6.4. FUNCTIONS OF CONTINUOUS RANDOM VARIABLES
155
Example 6.27. The Probability Integral Transform. Given a continuous random variable X with
strictly increasing CDF Fx, let the random variable Y be defined by Y - Fx(X). Then the distribu-
tion of Y is unif(min = 0, max = 1).
Proof. We employ the CDF method. First note that the support of Y is (0, 1). Then for any 0 < y <
1,
FY(y) = P (7 <>’) = P(Ey(X) < y).
Now since Fx is strictly increasing, it has a well defined inverse function F x 1 . Therefore,
PlExGO < y) = V(X < F- '(>■)) = Fx[F-'(y)] = y.
Summarizing, we have seen that FY(y) = y, 0 < y < 1. But this is exactly the CDF of a unif(min =
0, max =1) random variable. □
Fact 6.28. The Probability Integral Transform is true for all continuous random variables with
continuous CDFs, not just for those with strictly increasing CDFs ( but the proof is more compli-
cated). The transform is not true for discrete random variables, or for continuous random variables
having a discrete component (that is, with jumps in their CDF).
Example 6.29. Let Z ~ norm(mean = 0, sd = 1) and let U = Z2. What is the PDF of U1
Notice first that Z2 > 0, and thus the support of U is [0, oo). And for any u > 0,
F,j(u) = P (U < u) = P(Z2 < u).
But Z2 < u occurs if and only if - yfu < Z < \fit. The last probability above is simply the area
under the standard normal PDF from - sfu to sfit, and since <p is symmetric about 0, we have
P(Z2 < u) = 2 P(0 < Z < Vm) = 2 [Fz( Vm) - Fz(0)] = 20( fu)-\,
because 0(0) = 1 /2. To find the PDF of U we differentiate the CDF recalling that O' = (p.
fu(u) = (20( a/m) - lY = 2(f>( Vm) ■ = m_1/20( Vm).
x ’ 2 Vm
Substituting,
fu(u) = iF^2 e-^' ^2 = (2n u)~l^2e~u, u > 0.
V2tt
This is what we will later call a chi-square distribution with 1 degree of freedom. See Section 6.5.
6.4.3 How to do it with R
The distr package has functionality to investigate transformations of univariate distributions.
There are exact results for ordinary transformations of the standard distributions, and distr takes
advantage of these in many cases. For instance, the distr package can handle the transformation
in Example 6.23 quite nicely:
> library (distr )
> X <- Norm (mean - &, sd - 1)
>Y<-4-3*X
> Y
156
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
Distribution Object of Class: Norm
mean: 4
sd: 3
So distr “knows” that a linear transformation of a normal random variable is again normal,
and it even knows what the correct mean and sd should be. But it is impossible for distr to
know everything, and it is not long before we venture outside of the transformations that distr
recognizes. Let us try Example 6.22:
> Y <- exp (X)
> Y
Distribution Object of Class: AbscontDistribution
The result is an object of class AbscontDistribution, which is one of the classes that distr
uses to denote general distributions that it does not recognize (it turns out that Z has a lognormal
distribution; see Section 6.5). A simplified description of the process that distr undergoes when
it encounters a transformation Y = g(X) that it does not recognize is
1. Randomly generate many, many copies Aj, Xi, . . . , X„ from the distribution of X,
2. Compute ij = g(X\ ), IS = g(X 2), . . . , Y„ — g(X„) and store them for use.
3. Calculate the PDF, CDF, quantiles, and random variates using the simulated values of Y.
As long as the transformation is sufficiently nice, such as a linear transformation, the exponential,
absolute value, etc., the d-p-q functions are calculated analytically based on the d-p-q functions
associated with X. But if we try a crazy transformation then we are greeted by a warning:
> W <- sin(exp(X) + 27)
> W
Distribution Object of Class: AbscontDistribution
The warning confirms that the d-p-q functions are not calculated analytically, but are instead
based on the randomly simulated values of Y. We must be careful to remember this. The nature
of random simulation means that we can get different answers to the same question: watch what
happens when we compute \P(W < 0.5) using the W above, then define W again, and compute the
(supposedly) same IP( W < 0.5) a few moments later.
> p(W)(Q.5)
[1] 0.5793242
> W <- sin(exp(X) + 27)
> p(W) (&. 5)
[1] 0.5793242
The answers are not the same! Furthermore, if we were to repeat the process we would get yet
another answer for P (W < 0.5).
The answers were close, though. And the underlying randomly generated A’s were not the
same so it should hardly be a surprise that the calculated W’s were not the same, either. This serves
as a warning (in concert with the one that distr provides) that we should be careful to remember
that complicated transformations computed by R are only approximate and may fluctuate slightly
due to the nature of the way the estimates are calculated.
6.5. OTHER CONTINUOUS DISTRIBUTIONS
157
6.5 Other Continuous Distributions
6.5.1 Waiting Time Distributions
In some experiments, the random variable being measured is the time until a certain event occurs.
For example, a quality control specialist may be testing a manufactured product to see how long
it takes until it fails. An efficiency expert may be recording the customer traffic at a retail store to
streamline scheduling of staff.
The Exponential Distribution
We say that X has an exponential distribution and write X ~ expirate = A).
fx(x) = Ae~-lx, x > 0 (6.5.1)
The associated R functions are dexp(x, rate = 1), pexp, qexp, and rexp, which give the PDF,
CDF, quantile function, and simulate random variates, respectively.
The parameter A measures the rate of arrivals (to be described later) and must be positive. The
CDF is given by the formula
Fx(t) = 1 - e“'u, t > 0. (6.5.2)
The mean is p = 1/A and the variance is cr2 = 1/A1.
The exponential distribution is closely related to the Poisson distribution. If customers arrive
at a store according to a Poisson process with rate A and if Y counts the number of customers that
arrive in the time interval [0, f), then we saw in Section 5.6 that Y ~ pois(lambda = At). Now
consider a different question: let us start our clock at time 0 and stop the clock when the first
customer arrives. Let X be the length of this random time interval. Then X ~ exp(rate = A).
Observe the following string of equalities:
\P(X > t) = P(first arrival after time t ),
= P(no events in [0,0).
= P(T = 0),
= e~M,
where the last line is the PMF of Y evaluated at y = 0. In other words, Pi A < t) = 1 — e~'lf, which
is exactly the CDF of an exp(rate = A) distribution.
The exponential distribution is said to be memoryless because exponential random variables
"forget" how old they are at every instant. That is, the probability that we must wait an additional
five hours for a customer to arrive, given that we have already waited seven hours, is exactly the
probability that we needed to wait five hours for a customer in the first place. In mathematical
symbols, for any s, t > 0,
P(A > s + t\X > t) = P(A > s). (6.5.3)
See Exercise 6.5.
158
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
The Gamma Distribution
This is a generalization of the exponential distribution. We say that X has a gamma distribution and
write X ~ gammat shape = a, rate = d). It has PDF
fx(x) = xa~'e~Ax, X > 0. (6.5.4)
Frit)
The associated R functions are dgamma(x , shape, rate = 1), pgamraa, qgamma, and rgainma,
which give the PDF, CDF, quantile function, and simulate random variates, respectively. If a = 1
then X ~ exp(rate = A). The mean is /./ = a/ A and the variance is cr2 = a /A2.
To motivate the gamma distribution recall that if X measures the length of time until the first
event occurs in a Poisson process with rate A then X ~ exp(rate = A). If we let Y measure the
length of time until the alh event occurs then Y ~ gamma( shape = a, rate = A). When a is an
integer this distribution is also known as the Erlang distribution.
Example 6.30. At a car wash, two customers arrive per hour on the average. We decide to
measure how long it takes until the third customer arrives. If Y denotes this random time then
Y ~ gamma(shape = 3, rate = 1/2).
6.5.2 The Chi square, Student’s t, and Snedecor’s F Distributions
The Chi square Distribution
A random variable X with PDF
AW = r <6-5'5>
is said to have a chi-square distribution with p degrees of freedom. We write X ~ chisq(df = p ).
The associated R functions are dchisq(x, df), pchisq, qchisq, and rchisq, which give the
PDF, CDF, quantile function, and simulate random variates, respectively. See Figure 6.5.1. In an
obvious notation we may define Xo(p) as the number on the x-axis such that there is exactly a area
under the chisq(df = p) curve to its right.
The code to produce Figure 6.5.1 is
> curve (dchisq(x, df = 3), from - (9, to - 2(9, ylab = "y")
> ind <- c(4, 5, IIS, 15)
> for (i in ind) curve (dchisqCx, df = i) , (9, 2(9, add = TRUE)
Remark 6.31. Here are some useful things to know about the chi-square distribution.
1. If Z ~ norm(mean = 0, sd = 1), then Z2 ~ chisq(df = 1). We saw this in Example 6.29,
and the fact is important when it comes time to find the distribution of the sample variance,
S2. See Theorem 8.5 in Section 8.2.2.
2. The chi-square distribution is supported on the positive x-axis, with a right-skewed distribu-
tion.
3. The chisq(df = p) distribution is the same as a gamma) shape = p/2, rate = 1/2) distri-
bution.
6.5. OTHER CONTINUOUS DISTRIBUTIONS
159
LO
X
Figure 6.5.1: Chi square distribution for various degrees of freedom
160
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
4. The MGF of A ~ chisq(df = p) is
Mx(t) = (1 - 2 t)~p , t < 1/2. (6.5.6)
Student’s t distribution
A random variable X with PDF
fx(x) =
r [(r + l)/2]
V^Hr/2)
<r+ 1)/2
— oo < X < oo
(6.5.7)
is said to have Student’s t distribution with r degrees of freedom, and we write X ~ t(df = r). The
associated R functions are dt, pt, qt, and rt, which give the PDF, CDF, quantile function, and
simulate random variates, respectively. See Section 8.2.
Snedecor’s F distribution
A random variable X with p.d.f.
fx(x) =
T\{m + n)/2\ / m \m/1 ,„/2- 1 / , m \
r(m/2)r(u/2) U / X \+~n)
x > 0.
(6.5.8)
is said to have an F distribution with (in, n ) degrees of freedom. We write X ~ f(df 1 = m, df 2 = n).
The associated R functions are df(x, dfl, df2), pf, qf, and rf, which give the PDF, CDF,
quantile function, and simulate random variates, respectively. We define Fa(m, n) as the number on
the x-axis such that there is exactly a area under the f( df 1 = m, df2 = n) curve to its right.
Remark 6.32. Here are some notes about the F distribution.
1. If X ~ f(dfl = m, df2 = n) and Y = 1/A, then Y ~ f(dfl = n, df2 = m). Historically,
this fact was especially convenient. In the old days, statisticians used printed tables for their
statistical calculations. Since the F tables were symmetric in m and n, it meant that publishers
could cut the size of their printed tables in half. It plays less of a role today now that personal
computers are widespread.
2. If X ~ t(df = r), then X2 ~ f(df 1 = 1, df2 = r). We will see this again in Section 1 1.3.3.
6.5.3 Other Popular Distributions
The Cauchy Distribution
This is a special case of the Student’s t distribution. It has PDF
h(s) = h
l +
X - III
2T1
— oo < x < oo
(6.5.9)
We write A ~ cauchy(location = m, scale = fS). The associated R function is dcauchy(x,
location = 0, scale = 1).
It is easy to see that a cauchy(location = 0, scale = 1) distribution is the same as a t(df = 1)
distribution. The cauchy distribution looks like a norm distribution but with very heavy tails.
6.5. OTHER CONTINUOUS DISTRIBUTIONS
161
The mean (and variance) do not exist, that is, they are infinite. The median is represented by the
location parameter, and the scale parameter influences the spread of the distribution about its
median.
The Beta Distribution
This is a generalization of the continuous uniform distribution.
fx(x) =
T(a+P)
jx-l
(1 -xf-\ 0 < x < 1
(6.5.10)
r(a)TOS)
WewriteX ~ beta(shapel = a, shape2 = /?). The associated R function is dbeta(x, shapel,
shape2). The mean and variance are
R =
a
a +p
and cr- =
a/3
(a + py (a+/3+ 1)
(6.5.11)
See Example 6.3. This distribution comes up a lot in Bayesian statistics because it is a good model
for one’s prior beliefs about a population proportion p, 0 < p < 1.
The Logistic Distribution
fx(x) = - exp
cr
-oo < x < oo.
(6.5.12)
We write X ~ logisllocation = p, scale = cr). The associated R function is dlogis(x,
location = 0, scale = 1). The logistic distribution comes up in differential equations as a
model for population growth under certain assumptions. The mean is p and the variance is n2cr2 /3.
The Lognormal Distribution
This is a distribution derived from the normal distribution (hence the name). If U ~ norm(mean =
p, sd = cr), then X = e^has PDF
fx(x) =
: V27r
exp
-(lnx - p)1
2 cr2
0 < X < oo.
(6.5.13)
We write X ~ lnorm(meanlog = p, sdlog = cr). The associated R function is dlnorm(x,
meanlog = 0, sdlog = 1). Notice that the support is concentrated on the positive x axis; the
distribution is right-skewed with a heavy tail. See Example 6.22.
The Weibull Distribution
This has PDF
Mx) = | (j) exP , x>0. (6.5.14)
We write X ~ weibulKshape = a, scale = f3). The associated R function is dweibull(x,
shape, scale = 1).
162
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
6.5.4 How to do it with R
There is some support of moments and moment generating functions for some continuous prob-
ability distributions included in the actuar package [25]. The convention is m in front of the
distribution name for raw moments, and mgf in front of the distribution name for the moment gen-
erating function. At the time of this writing, the following distributions are supported: gamma,
inverse Gaussian, (non-central) chi-squared, exponential, and uniform.
Example 6.33. Calculate the hist four raw moments for X ~ gamma! shape = 13, rate = 1) and
plot the moment generating function.
We load the actuar package and use the functions ragamma and mgfgamma:
> library(actuar)
> mgamma(l:4, shape = 13, rate = 1)
[1] 13 182 273® 4368®
For the plot we can use the function in the following form:
> plot(function(x) {
+ mgfgamma(x, shape = 13, rate = 1)
+ }, from = -6.1, to = 8.1, ylab = "gamma mgf")
gamma mgf
6.5. OTHER CONTINUOUS DISTRIBUTIONS
163
X
Figure 6.5.2: Plot of the gamma! shape = 13, rate = 1) MGF
164
CHAPTER 6. CONTINUOUS DISTRIBUTIONS
Chapter Exercises
Exercise 6.1. Find the constant c so that the given function is a valid PDF of a random variable X.
1. f(x) = Cxn, 0 < x < 1 .
2. fix) = Cxe *, 0 < x < oo.
3. f{x) = , 7 < x < oo.
4. f(x) = Cx\ 1 - x)2, 0 < x < 1.
5. fix) = Ci 1 + x2/4)~‘, -oo < x < oo.
Exercise 6.2. For the following random experiments, decide what the distribution of X should be.
In nearly every case, there are additional assumptions that should be made for the distribution to
apply; identify those assumptions (which may or may not strictly hold in practice).
1. We throw a dart at a dart board. Let X denote the squared linear distance from the bulls-eye
to the where the dart landed.
2. We randomly choose a textbook from the shelf at the bookstore and let P denote the propor-
tion of the total pages of the book devoted to exercises.
3. We measure the time it takes for the water to completely drain out of the kitchen sink.
4. We randomly sample strangers at the grocery store and ask them how long it will take them
to drive home.
Exercise 6.3. If Z is norm(mean = 0, sd = 1), find
1. P(Z > 2.64)
> pnorm(2.64, lower . tail = FALSE)
[1] 8. 004145301
2. P(0 < Z < 0.87)
> pnormiQ . 87) - 1/2
[1] 8.3878498
3. P(|Z| > 1.39) (Hint: draw a picture!)
> 2 * pnorm(-l . 39)
[1] 8.1645289
Exercise 6.4. Calculate the variance of X ~ unif(min = a, max = b). Hint: First calculate IE X2.
type the exercise here
Exercise 6.5. Prove the memoryless property for exponential random variables. That is, for X ~
exp(rate = A) show that for any s, t > 0,
TPiX> s + t\X>t) = W>iX> s).
Chapter 7
Multivariate Distributions
We have built up quite a catalogue of distributions, discrete and continuous. They were all uni-
variate, however, meaning that we only considered one random variable at a time. We can imagine
nevertheless many random variables associated with a single person: their height, their weight,
their wrist circumference (all continuous), or their eye/hair color, shoe size, whether they are right
handed, left handed, or ambidextrous (all categorical), and we can even surmise reasonable proba-
bility distributions to associate with each of these variables.
But there is a difference: for a single person, these variables are related. For instance, a person’s
height betrays a lot of information about that person’s weight.
The concept we are hinting at is the notion of dependence between random variables. It is the
focus of this chapter to study this concept in some detail. Along the way, we will pick up additional
models to add to our catalogue. Moreover, we will study certain classes of dependence, and clarify
the special case when there is no dependence, namely, independence.
The interested reader who would like to learn more about any of the below mentioned multi-
variate distributions should take a look at Discrete Multivariate Distributions by Johnson el al [49]
or Continuous Multivariate Distributions [54] by Kotz et al.
What do I want them to know?
• the basic notion of dependence and how it is manifested with multiple variables (two, in
particular)
• joint versus marginal distributions/expectation (discrete and continuous)
• some numeric measures of dependence
• conditional distributions, in the context of independence and exchangeability
• some details of at least one multivariate model (discrete and continuous)
• what it looks like when there are more than two random variables present
165
166
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
7.1 Joint and Marginal Probability Distributions
Consider two discrete random variables X and Y with PMFs fx and fY that are supported on the
sample spaces S x and S Y, respectively. Let S x. y denote the set of all possible observed pairs (x, yj,
called the joint support set of X and Y. Then the joint probability mass function of X and Y is the
function fXY defined by
fx,y(x,y) = P(X = x, Y = y), for (x,y) e SXJ.
(7.1.1)
Every joint PMF satisfies
fxj(x,y ) > 0 for all (x,y) e SXY,
(7.1.2)
and
X fxj(x, y) = 1.
(7.1.3)
( x,y)eSx,r
It is customary to extend the function fXY to be defined on all of R2 by setting fXj(x,y) = 0 for
(x,y) i SXJ.
In the context of this chapter, the PMFs fx and fy are called the marginal PMFs of X and Y,
respectively. If we are given only the joint PMF then we may recover each of the marginal PMFs
by using the Theorem of Total Probability (see Equation4.4.5): observe
fx(x) = IP(X = x),
yeS y
= ^ fxj(x,y).
yeS y
By interchanging the roles of X and Y it is clear that
fy(y) = X fxAxA). (7.1.7)
xeSy
Given the joint PMF we may recover the marginal PMFs, but the converse is not true. Even if we
have both marginal distributions they are not sufficient to determine the joint PMF; more informa-
tion is needed1 .
Associated with the joint PMF is the joint cumulative distribution function FXY defined by
Fx,Y(x,y) = P(X < x, Y < y), for ( x,y ) 6 K2.
The bivariate joint CDF is not quite as tractable as the univariate CDFs, but in principle we could
calculate it by adding up quantities of the form in Equation 7.1.1. The joint CDF is typically not
used in practice due to its inconvenient form; one can usually get by with the joint PMF alone.
We now introduce some examples of bivariate discrete distributions. The first we have seen
before, and the second is based on the first.
*We are not at a total loss, however. There are Frechet bounds which pose limits on how large (and small) the joint
distribution must be at each point.
(7.1.4)
(7.1.5)
(7.1.6)
7. 1 . JOINT AND MARGINAL PROBABILITY DISTRIB UTIONS
167
Example 7.1. Roll a fair die twice. Let X be the face shown on the first roll, and let Y be the face
shown on the second roll. We have already seen this example in Chapter 4, Example 4.30. For this
example, it suffices to define
h.,U.y) = 5^.
The marginal PMFs are given by fx Or) =
since
6
fx(x) = Yj
y=\
x — 1, . . . , 6, y = 1, . . . ,6.
1/6, x — 1,2,..., 6, and fy(y)
1/6, y = 1,2,. ..,6,
and the same computation with the letters switched works for Y.
In the previous example, and in many other ones, the joint support can be written as a product
set of the support of X “times” the support of Y, that is, it may be represented as a cartesian product
set, or rectangle, Sx.y = Sx x Sy, where Sx X Sy = {(x,y) : x e S x, y e Sy). As we shall see
presently in Section 7.4, this form is a necessary condition for X and Y to be independent (or
alternatively exchangeable when S x = Sy). But please note that in general it is not required for
S x,y to be of rectangle form. We next investigate just such an example.
Example 7.2. Let the random experiment again be to roll a fair die twice, except now let us define
the random variables U and V by
U = the maximum of the two rolls, and
V = the sum of the two rolls.
We see that the support of U is S u = {1,2, . . . , 6} and the support of V is Sy = {2, 3, . . . , 12). We
may represent the sample space with a matrix, and for each entry in the matrix we may calculate
the value that U assumes. The result is in the left half of Table 7.1.
We can use the table to calculate the marginal PMF of U, because from Example 4.30 we know
that each entry in the matrix has probability 1 /36 associated with it. For instance, there is only one
outcome in the matrix with U = 1, namely, the top left corner. This single entry has probability
1/36, therefore, it must be that fu( 1) = IP(6' = 1) = 1/36. Similarly we see that there are three
entries in the matrix with U - 2, thus fu( 2) = 3/36. Continuing in this fashion we will find the
marginal distribution of U may be written
/£/(«} m = 1, 2, ... ,6. (7.1.8)
36
We may do a similar thing for V; see the right half of Table 7.1. Collecting all of the probability
we will find that the marginal PMF of V is
6 — |v — 71
fv(y)= ^7— V = 2, 3,..., 12. (7.1.9)
36
We may collapse the two matrices from Table 7.1 into one, big matrix of pairs of values ( u , v).
The result is shown in Table 7.2.
Again, each of these pairs has probability 1/36 associated with it and we are looking at the joint
PDF of (U, V) albeit in an unusual form. Many of the pairs are repeated, but some of them are not:
168
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
u
1
2
3
4
5
6
V
1
2
3
4
5
6
1
1
2
3
4
5
6
1
2
3
4
5
6
7
2
2
2
3
4
5
6
2
3
4
5
6
1
8
3
3
3
3
4
5
6
3
4
5
6
7
8
9
4
4
4
4
4
5
6
4
5
6
7
8
9
10
5
5
5
5
5
5
6
5
6
7
8
9
10
11
6
6
6
6
6
6
6
6
7
8
9
10
11
12
(a) U -
= max(X, V)
(b) V =
X+Y
Table 7.1: Maximum U and sum V of a pair of dice rolls ( X , Y)
(U,V)
1
2
3
4
5
6
1
(1,2)
(2,3)
(3,4)
(4.5)
(5,6)
(6,7)
2
(2,3)
(2,4)
(3,5)
(4,6)
(5,7)
(6,8)
3
(3,4)
(3,5)
(3,6)
(4,7)
(5,8)
(6,9)
4
(4,5)
(4,6)
(4,7)
(4,8)
(5,9)
(6,10)
5
(5,6)
(5,7)
(5,8)
(5,9)
(5,10)
(6,11)
6
(6,7)
(6,8)
(6,9)
(6,10)
(6,11)
(6,12)
Table 7.2: Joint values of U - max(X, Y) and V — X + Y
2 3 4 5 6 7 8 9 10 11 12
Total
1
1/36
1/36
2
2/36 1/36
3/36
3
2/36 2/36 1/36
5/36
4
2/36 2/36 2/36 1/36
7/36
5
2/36 2/36 2/36 2/36 1/36
9/36
6
2/36 2/36 2/36 2/36 2/36 1/36
11/36
Total
1/36 2/36 3/36 4/36 5/36 6/36 5/36 4/36 3/36 2/36 1/36
1
Table 7.3: The joint PMF of ( U, V )
The outcomes of U are along the left and the outcomes of V are along the top. Empty entries in the table have
zero probability. The row totals (on the right) and column totals (on the bottom) correspond to the marginal
distribution of U and V, respectively.
7. 1 . JOINT AND MARGINAL PROBABILITY DISTRIB UTIONS
169
(1,2) appears twice, but (2, 3) appears only once. We can make more sense out of this by writing a
new table with U on one side and V along the top. We will accumulate the probability just like we
did in Example 7.1. See Table 7.3.
The joint support of (U,V) is concentrated along the main diagonal; note that the nonzero
entries do not form a rectangle. Also notice that if we form row and column totals we are doing
exactly the same thing as Equation 7.1.7, so that the marginal distribution of U is the list of totals
in the right “margin” of the Table 7.3, and the marginal distribution of V is the list of totals in the
bottom “margin”.
Continuing the reasoning for the discrete case, given two continuous random variables X and
Y there similarly exists2 a function fxy(x,y) associated with X and Y called the joint probability
density function of X and Y . Every joint PDF satisfies
In the continuous case there is not such a simple interpretation for the joint PDF; however, we
do have one for the joint CDF, namely.
for (x, y) e R2. If X and Y have the joint PDF fx y, then the marginal density of X may be recovered
by
fx(x) = f fxj(x, y) dv, xeXr (7.1.12)
JSy
and the marginal PDF of Y may be found with
fr(y) = f fxAx,y) dx, yeSF. (7.1.13)
JSX
Example 7.3. Let the joint PDF of (X, Y) be given by
fxj(x,y) > 0 for all (x,y) e SXJ,
(7.1.10)
and
(7.1.11)
The marginal PDF of X is
^Strictly speaking, the joint density function does not necessarily exist. But the joint CDF always exists.
170
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
for 0 < x < 1, and the marginal PDF of Y is
fr(y) =
for 0 < y < 1. In this example the joint support set was a rectangle [0, 1] x [0,1], but it turns out
that X and Y are not independent. See Section 7.4.
7.1.1 How to do it with R
We will show how to do Example 7.2 using R; it is much simpler to do it with R than without. First
we set up the sample space with the rolldie function. Next, we add random variables U and V
with the addrv function. We take a look at the very top of the data frame (probability space) to
make sure that everything is operating according to plan.
> S <- rolldie(2 , makespace = TRUE)
> S <- addrv(S, FUN = max, invars = c("Xl", "X 2"), name = "U")
> S <- addrv(S, FUN = sum, invars = c("Xl", "X 2"), name = "V")
> head(S)
XI X2 U V probs
11112 8.02777778
22123 8.82777778
33134 8.82777778
44145 8.82777778
55156 8.82777778
66167 8.82777778
Yes, the U and V columns have been added to the data frame and have been computed correctly.
This result would be fine as it is, but the data frame has too many rows: there are repeated pairs
(u,v) which show up as repeated rows in the data frame. The goal is to aggregate the rows of S
such that the result has exactly one row for each unique pair (u, v) with positive probability. This
sort of thing is exactly the task for which the marginal function was designed. We may take a
look at the joint distribution of U and V (we only show the first few rows of the data frame, but the
complete one has 1 1 rows).
> UV <- marginal (S, vars = c("U", "V"))
> head(UV)
U V probs
1 1 2 8.82777778
223 8.85555556
324 8.82777778
434 8.85555556
7. 1 . JOINT AND MARGINAL PROBABILITY DISTRIB UTIONS
171
5 B 5 0.05555556
6 4 5 0.05555556
The data frame is difficult to understand. It would be better to have a tabular display like Table
7.3. We can do that with the xtabs function.
> xtabs (round (probs , 3) ~ U + V, data = UV)
V
u
2
3
4
5
6
7
8
9
10
11
12
1
0.
.028
0.
.000
0.
.000
0.
.000
0,
.000
0.
.000
0.
.000
0.
.000
0.
.000
0.
.000
0.
.000
2
0.
.000
0.
.056
0.
.028
0.
.000
0,
.000
0.
.000
0.
.000
0.
.000
0.
.000
0.
.000
0.
.000
3
0.
.000
0.
.000
0.
.056
0.
.056
0,
.028
0.
.000
0.
.000
0.
.000
0.
.000
0.
.000
0.
.000
4
0.
.000
0.
.000
0.
.000
0.
.056
0,
.056
0.
.056
0.
.028
0.
.000
0.
.000
0.
.000
0.
.000
5
0.
.000
0.
.000
0.
.000
0.
.000
0,
.056
0.
.056
0.
.056
0.
.056
0.
.028
0.
.000
0.
.000
6
0.
.000
0.
.000
0.
.000
0,
.000
0.
.000
0.
.056
0.
.056
0.
.056
0.
.056
0.
.056
0.
.028
Compare these values to the ones shown in Table 7.3. We can repeat the process with marginal
to get the univariate marginal distributions of U and V separately.
> marginal (UV, vars = "U")
U probs
1 1 0.02777778
2 2 0 . 0833BBB3
3 3 0.13888889
4 4 0.19444444
5 5 0.25000000
6 6 0.30555556
> head (marginal (UV, vars = "V"))
V probs
1 2 0.02777778
2 3 0.05555556
3 4 0.08333333
4 5 0.11111111
5 6 0.13888889
6 7 0.16666667
Another way to do the same thing is with the rowSums and colSums of the xtabs object.
Compare
> temp <- xtabs (probs ~ U + V, data = UV)
> rowSums (temp)
1 2 3 4 5 6
0.0 2777778 0.08333333 0.13888889 0.19444444 0.25000000 0.30555556
> colSums(temp)
172
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
2 B 4 5 6 7
0 . 02777778 0.05555556 0.08333333 0.11111111 0.13888889 0.16666667
8 9 10 11 12
0.13888889 0.11111111 0.08333333 0.05555556 0. 02777778
You should check that the answers that we have obtained exactly match the same (somewhat
laborious) calculations that we completed in Example 7.2.
7.2 Joint and Marginal Expectation
Given a function g with arguments (x,y) we would like to know the long-run average behavior
of g(X, Y) and how to mathematically calculate it. Expectation in this context is computed in the
pedestrian way. We simply integrate (sum) with respect to the joint probability density (mass)
function.
or in the discrete case
E g(X, Y ) =
g(x,y) fxAvy) dxdy,
Sx,Y
E g(X, Y) = ZZ g(x,y) fx.y(x,y).
( x,y)eSx,r
(7.2.1)
(7.2.2)
7.2.1 Covariance and Correlation
There are two very special cases of joint expectation: the covariance and the correlation. These
are measures which help us quantify the dependence between X and Y .
Definition 7.4. The covariance of X and Y is
Cov(X, Y) = E(X - IE X)(Y - IE Y).
(7.2.3)
By the way, there is a shortcut formula for covariance which is almost as handy as the shortcut
for the variance:
Cov(Y, Y) = IE (XY) - (IE X)(1E Y). (7.2.4)
The proof is left to Exercise 7.1.
The Pearson product moment correlation between X and Y is the covariance between X and Y
rescaled to fall in the interval [-1, 1], It is formally defined by
Corr(Y, Y) =
Cov(X, Y)
o-xo-Y
(7.2.5)
The correlation is usually denoted by pXY or simply p if the random variables are clear from context.
There are some important facts about the correlation coefficient:
1. The range of correlation is -1 < pXY < 1.
2. Equality holds above (px,Y = +1) if and only if Y is a linear function of X with probability
one.
7.2. JOINT AND MARGINAL EXPECTATION
173
Example 7.5. We will compute the covariance for the discrete distribution in Example 7.2. The
expected value of U is
1EU = ^ ufuiu ) = ^
U= 1 M=1
and the expected value of V is
12
2u-l
36
= l|il + 2|^-| +
6 - |7 - v[
36
= 2
36
+ 3
36
-6|11U 161
+ 12
36/ 36 ’
= 7,
and the expected value of UV is
6 12
Ef/y = ££Mv/^,v) = 1 '2(^j + 2-3^j + ■■ ■ + 6 ■ 12
M=1 v=2
Therefore the covariance of (U, V) is
308
~9~
308 161 35
Cov(t/, F) = E UV - (E t/) (E F) = — — • 7 = — .
y 36 12
All we need now are the standard deviations of U and V to calculate the correlation coefficient
(omitted).
We will do a continuous example so that you can see how it works.
Example 7.6. Let us find the covariance of the variables ( X , Y ) from Example 7.3. The expected
value of X is
-f
1
EX = I x - xh — dx= -xJ H — x'
1
r=0
and the expected value of Y is
Ef =
Finally, the expected value of XY is
EAT =
Li
jli + ZK'. A/+
y=0
5’
20’
X X xyj(x+y‘)dxdy-
f
2 3 3 4
5x>,+ !o^
r1 (2 3
= i\5y+^
/ dy.
1
5 + 50’
which is 13/50. Therefore the covariance of ( X , T) is
1
Too'
174
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
7.2.2 How to do it with R
There are not any specific functions in the prob package designed for multivariate expectation.
This is not a problem, though, because it is easy enough to do expectation the long way - with
column operations. We just need to keep the definition in mind. For instance, we may compute the
covariance of ((/, V) from Example 7.5.
> Eu <- sum(S$U * SSprobs)
> Ev <- sum(S$V * SSprobs)
> Euv <- sum(S$U * S$V * SSprobs)
> Euv - Eu * Ev
[1] 2.916667
Compare this answer to what we got in Example 7.5.
To do the continuous case we could use the computer algebra utilities of Yacas and the asso-
ciated R package Ryacas [35]. See Section 7.7.1 for another example where the Ryacas package
7.3 Conditional Distributions
If x G S x is such that fx(x) > 0, then we define the conditional density of Y\ X = x, denoted fY\x,
by
Example 7.8. Let the joint PDF of X and Y be given by
Bayesian Connection
Conditional distributions play a fundamental role in Bayesian probability and statistics. There is
a parameter 8 which is of primary interest, and about which we would like to learn. But rather
than observing 8 directly, we instead observe a random variable X whose probability distribution
depends on 8. Using the information we provided by X, we would like to update the information
that we have about 8.
Our initial beliefs about 8 are represented by a probability distribution, called the prior distri-
bution, denoted by it. The PDF fx\e is called the likelihood function, also called the likelihood of
X conditional on 8. Given an observation X = x, we would like to update our beliefs n to a new
distribution, called the posterior distribution of 8 given the obseri’ation X = x, denoted n^x. It
may seem a mystery how to obtain nB\x based only on the information provided by n and fx \g, but it
should not be. We have already studied this in Section 4.8 where it was called Bayes’ Rule:
appears.
(7.3.1)
We define fx\y in a similar fashion.
Example 7.7. Let the joint PMF of X and Y be given by
fxj(x,y) =
(7.3.2)
7.3. CONDITIONAL DISTRIBUTIONS
175
Compare the above expression to Equation 4.8.1.
Example 7.9. Suppose the parameter 9 is the P(Heads) for a biased coin. It could be any value
from 0 to 1. Perhaps we have some prior information about this coin, for example, maybe we
have seen this coin before and we have reason to believe that it shows Heads less than half of the
time. Suppose that we represent our beliefs about 9 with a beta(shapel = 1, shape2 = 3) prior
distribution, that is, we assume
9 ~ n{9) = 3(1— 9)2, 0<G<\.
To learn more about 9, we will do what is natural: flip the coin. We will observe a random variable X
which takes the value 1 if the coin shows Heads, and 0 if the coin shows Tails. Under these circum-
stances, X will have a Bernoulli distribution, and in particular, X\ 9 ~ binom(size = 1, prob = 9):
fxw(x\6) = 0X(\ x = 0,1.
Based on the observation X = x, we will update the prior distribution to the posterior distribution,
and we will do so with Bayes’ Rule: it says
n(9\x) oc n{9)f(x\9),
= 0*(1 - 9)l~x • 3(1 - 9)2,
= 3 9x(l - 9)3~x, 0<9<1,
where the constant of proportionality is given by
3 ux( 1 - u)3~xdu = [3 u(1+xyl(l - u)(4~xyidu = 3 r<1 + Y>n4 - r) ,
J T[(l + x) + (4 - x)]
the integral being calculated by inspection of the formula for a beta(shapel = 1 + x, shape2 =
4 - x) distribution. That is to say, our posterior distribution is precisely
9\x ~ beta! shape 1 = 1 + x, shape2 =4- x).
The Bayesian statistician uses the posterior distribution for all matters concerning inference
about 9.
Remark 7.10. We usually do not restrict ourselves to the observation of only one X conditional
on 9. In fact, it is common to observe an entire sample X\, X2,. . . ,Xn conditional on 9 (which
itself is often multidimensional). Do not be frightened, however, because the intuition is the
same. There is a prior distribution n{9), a likelihood f(x\ ,x2,..., x„\9), and a posterior distribu-
tion n(0\xi , X2, ■ ■ ■ , xn). Bayes’ Rule states that the relationship between the three is
n(9\xi ,x2,..., xn ) oc n(9) f(x 1 ,x2,..., xn\ 0),
where the constant of proportionality is J n(u) f(x i,x2, ■ ■ ■ , xn\u) du. Any good textbook on Bayesian
Statistics will explain these notions in detail; to the interested reader I recommend Gelman [33] or
Lee [57],
176
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
7.4 Independent Random Variables
7.4.1 Independent Random Variables
We recall from Chapter 4 that the events A and B are said to be independent when
P (A C\B) = P(A) P(fi).
If it happens that
P(X = x,Y = y) = P(X = x ) P(F = y), for every x e Sx, y e SY,
(7.4.1)
(7.4.2)
then we say that X and Y are independent random variables. Otherwise, we say that X and Y are
dependent. Using the PMF notation from above, we see that independent discrete random variables
satisfy
fx.viX y) = fx(x)fY(y) for every x € Sx, y e SY. (7.4.3)
Continuing the reasoning, given two continuous random variables X and Y with joint PDF fx.Y and
respective marginal PDFs fx and fY that are supported on the sets S x and S Y, if it happens that
fxAx, y) = fx(x)fY(y) for every x e Sx, y e SY, (7.4.4)
then we say that X and Y are independent.
Example 7.11. In Example 7.1 we considered the random experiment of rolling a fair die twice.
There we found the joint PMF to be
fxjivy) = x = 1,. ..,6, y = 1,...,6,
36
and we found the marginal PMFs fx(x ) = 1/6. x = 1 , 2, .... 6, and fY(y) = 1/6 , y — 1 , 2, . . . , 6.
Therefore in this experiment X and Y are independent since for every x and y in the joint support
the joint PMF satisfies
Example 7.12. In Example 7.2 we considered the same experiment but different random variables
U and V. We can prove that U and V are not independent if we can find a single pair (m, v) where
the independence equality does not hold. There are many such pairs. One of them is (6, 12):
/Rv(6,12) = 4*(1I)(_L) = /„(6)/v(12).
Independent random variables are very useful to the mathematician. They have many, many,
tractable properties. We mention some of the more important ones.
Proposition 7.13. IfX and Y are independent, then for any functions u and v,
IE {u(X)v( Y)) = (IE u(X)) (IE v(F)) .
(7.4.5)
7.4. INDEPENDENT RANDOM VARIABLES
111
Proof. This is straightforward from the definition.
E(M(X)v(T))
If
If
J u(x)fx(x)dx J' v(y)fr(y)dy
u(x)v(y) fXj(x, y) dxdv
u(x)v(y) fx(x) fY(y ) dxdv
and this last quantity is exactly (E u(X )) (E v(Y )). □
Now that we have Proposition 7.13 we mention a corollary that will help us later to quickly
identify those random variables which are not independent.
Corollary 7.14. If X and Y are independent, then Cov(X, Y ) = 0, and consequently, Corr(X, Y) = 0.
Proof When X and Y are independent then E XY = EXE Y. And when the covariance is zero
the numerator of the correlation is 0. □
Remark 7.15. Unfortunately, the converse of Corollary 7.14 is not true. That is, there are many
random variables which are dependent yet their covariance and correlation is zero. For more details,
see Casella and Berger [13],
Proposition 7.13 is useful to us and we will receive mileage out of it, but there is another fact
which will play an even more important role. Unfortunately, the proof is beyond the techniques
presented here. The inquisitive reader should consult Casella and Berger [13], Resnick [70], etc.
Fact 7.16. IfX and Y are independent, then u(X) and v(F) are independent for any functions u and
v.
7.4.2 Combining Independent Random Variables
Another important corollary of Proposition 7.13 will allow us to find the distribution of sums of
random variables.
Corollary 7.17. If X and Y are independent, then the moment generating function ofX + Y is
Mx+Y(t) = Mx(t) ■ Mr(t). (7.4.6)
Proof. Choose u(x) = ex and v(y) = e-’ in Proposition 7.13, and remember the identity e'IA+v) =
etx e'-v. □
Let us take a look at some examples of the corollary in action.
Example 7.18. Let X ~ binom(size = m, prob = p) and Y ~ binom(size = m, prob = p) be
independent. Then X + Y has MGF
Mx+Y(t) = Mx(t) MY(t ) = {q + pe')"' (q + pe')"1 = (q + pe')"’+"2 ,
which is the MGF of a binom(size = n\ + 112, prob = p) distribution. Therefore, X + Y ~
binom(size = m + ni, prob = p).
178
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
Example 7.19. Let X ~ norm(mean = /zi, sd = cr\) and Y ~ norm(mean = ^2, sd = cr2) be
independent. Then X + Y has MGF
Mx(t) M Y(t) = exp (yujf + t2 0-^/2} exp {p2t + r<x2/2} = exp {(//! +, p2)t + r ( cr 2 + of) /2} ,
which is the MGF of a norm(mean = p\ + p2, sd = y<x2 + of) distribution.
Even when we cannot use the MGF trick to identify the exact distribution of a linear combina-
tion of random variables, we can still say something about its mean and variance.
Proposition 7.20. Let X\ and X2 be independent with respective population means p \ and p2 and
population standard deviations cr\ and cr2. For given constants a 1 and a2, define Y — a \X\ + a2X2.
Then the mean and standard deviation ofY are given by the formulas
t 2 2 2 2\F2
pY = a\p\ + a2p2, o-y = (a1(Tl + a2(T2)
(7.4.7)
Proof. We use Proposition 5.11:
IE Y — IE (a iX\ + a2X2) — a\ TEX\ + a2 EX2 = ci\P\ + a2p2-
For the standard deviation, we will find the variance and take the square root at the end. And
to calculate the variance we will first compute IE Y 2 with an eye toward using the identity cr2. =
IE Y2 - (IE Y)2 as a final step.
IE Y2 = IE (aiXi + a2X2)2 = IE [a]X 2 + a22Xj + 2aia2XiX2) .
Using linearity of expectation the IE distributes through the sum. Now IE A2 = cr2 + p2. for i = 1
and 2 and IE X\X2 — IE X\ IE X2 = p\p2 because of independence. Thus
IET2 = a\{cr\ + p\) + a\(o\+ p\) + 2aia2pxp2,
= ajcr2 + at,cr\ + (a~iP] + a2p2 + 7.a\a2p\p2) .
But notice that the expression in the parentheses is exactly {a\p\ + a2p2)2 = (IE Y)2, so the proof is
complete.. □
7.5 Exchangeable Random Variables
Two random variables X and Y are said to be exchangeable if their joint CDF is a symmetric
function of its arguments:
FXj(x,y) = Fxj(y,x), for all (x, y) e R2. (7.5.1)
When the joint density / exists, we may equivalently say that X and Y are exchangeable if f(x,y) =
f(y, x) for all (x,y).
Exchangeable random variables exhibit symmetry in the sense that a person may exchange one
variable for the other with no substantive changes to their joint random behavior. While indepen-
dence speaks to a lack of influence between the two variables, exchangeability aims to capture the
symmetry between them.
7. 6. THE BIVARIATE NORMAL DISTRIB UTION
179
Example 7.21. Let X and Y have joint PDF
fxj(x,y ) = (1 + a)A2e~A(x+y) + a(2 A)2e-2A(x+y) - 2aA2 (e-A(2x+y) + e~A(x+2y)) . (7.5.2)
It is straightforward and tedious to check that JJ f = 1. We may see immediately that fXj(x,y ) =
fx.y(y,x ) for all (x,y), which confirms that X and Y are exchangeable. Here, a is said to be an
association parameter. This particular example is one from the Farlie-Gumbel-Morgenstern family
of distributions; see [54].
Example 7.22. Suppose X and Y are i.i.d. binom(size = n, prob = p ). Then their joint PMF is
fxj(x,y) = fx(x)fY(y)
= pxo -pT~y,
and the value is the same if we exchange x and y. Therefore ( X , Y) are exchangeable.
Looking at Example 7.22 more closely we see that the fact that (X, Y ) are exchangeable has
nothing to do with the binom(size = n, prob = p) distribution; it only matters that they are
independent (so that the joint PDF factors) and they are identically distributed (in which case we
may swap letters to no effect). We could just have easily used any other marginal distribution. We
will take this as a proof of the following proposition.
Proposition 7.23. IfX and Y are i.i.d. (with common marginal distribution F) then X and Y are
exchangeable.
Exchangeability thus contains i.i.d. as a special case.
7.6 The Bivariate Normal Distribution
The bivariate normal PDF is given by the unwieldy formula
fxj(x,y )
1
exp <
1
2n ctxcty yj\ - p1 ( 2(1 p~)
x~px
ox
+ 2 p
ox
x- px\(y - py\ (y - py
<ty
O" Y
for (x, >j e R2. We write (X, Y) ~ mvnorm(mean = p, sigma = £), where
p = (px, pY)T, ^
cr2 po-xcry
po-xo-y O-2
(7.6.1)
(7.6.2)
See Appendix E. The vector notation allows for a more compact rendering of the joint PDF:
fxj(x) -
1
2tt|S|1/2
1
exp ^ - T (x - P)T S_1 (x - p) L
(7.6.3)
180
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
where in an abuse of notation we have written x for (x,y). Note that the formula only holds when
p 4 ±1.
Remark 7 .24. In Remark 7. 15 we noted that just because random variables are uncorrelated it does
not necessarily mean that they are independent. However, there is an important exception to this
rule: the bivariate normal distribution. Indeed, ( X , Y ) ~ mvnorm(mean = p, sigma = 2) are
independent if and only if p — 0.
Remark 7.25. Inspection of the joint PDF shows that if px = Py and crx = cry then X and Y are
exchangeable.
The bivariate normal MGF is
MXj(X) = exp (pTt + ^tT2tj , (7.6.4)
where t = (t\,t2).
The bivariate normal distribution may be intimidating at first but it turns out to be very tractable
compared to other multivariate distributions. An example of this is the following fact about the
marginals.
Fact 7.26. If(X, Y) ~ mvnormfmean = p, sigma = 2) then
X ~ normfmean = px, sd = crx) and Y ~ normfmean = py, sd = cry). (7.6.5)
From this we immediately get that E X — px and Var(A) = <x3 (and the same is true for Y with
the letters switched). And it should be no surprise that the correlation between X and Y is exactly
Coit(A, Y ) = p.
Proposition 7.27. The conditional distribution ofY \ X — x is norm(mean = pY\x, sd = crY\x), where
PY\x
CTy
pY + p — (x - px) , and o-y\x
crx
'■ cry
V
1-P2.
(7.6.6)
There are a few things to note about Proposition 7.27 which will be important in Chapter 11.
First, the conditional mean of Y\x is linear in x, with slope
cry
P — •
crx
(7.6.7)
Second, the conditional variance of Y\x is independent of x.
7.6.1 How to do it with R
The multivariate normal distribution is implemented in both the mvtnorm package [34] and the
mnormt package [17]. We use the mvtnorm package in this book simply because it is a dependency
of another package used in the book.
The mvtnorm package has functions dmvnorm and rmvnorm for the PDF and to generate ran-
dom vectors, respectively. Let us get started with a graph of the bivariate normal PDF. We can
make the plot with the following code3, where the workhorse is the persp function in base R.
'Another way to do this is with the curve3d function in the emdbook package [9], It looks like this:
7.7. BIVARIATE TRANSFORMATIONS OF RANDOM VARIABLES
181
> library(mvtnorm)
> x <- y <- seq(from = -3, to = 3, length. out = 36)
> f <- function(x, y) dmvnorm(cbind(x , y) , mean = c (6, 6),
+ sigma = diag(2))
> z <- outer(x, y, FUN = f)
> persp(x, y, z, theta = -36, phi = 36, ticktype = "detailed")
We chose the standard bivariate normal, mvnorm(mean = 0, sigma = I), to display.
7.7 Bivariate Transformations of Random Variables
We studied in Section 6.4 how to find the PDF of Y = g(X) given the PDF of X. But now we have
two random variables X and Y, with joint PDF fxj, and we would like to consider the joint PDF of
two new random variables
U = g(X, Y) and V = h(X, Y), (7.7.1)
where g and h are two given functions, typically “nice” in the sense of Appendix E.6.
Suppose that the transformation (x,y) i — * (u, v) is one-to-one. Then an inverse transformation
x = x(u, v) and v = y(u, v) exists, so let d(x,y)/d(u, v) denote the Jacobian of the inverse transfor-
mation. Then the joint PDF of ((/, V) is given by
fu.viu, v) = fx.v [x(u, v), y(u, v)]
d(x,y)
d(u, v)
or we can rewrite more shortly as
fu,v(u,v) = fxj(x, y)
d(x,y )
d(u, v)
(7.7.2)
(7.7.3)
Take a moment and compare Equation 7.7.3 to Equation 6.4.2. Do you see the connection?
Remark 7.28. It is sometimes easier to postpone solving for the inverse transformation x = x(u, v)
and y = y(u, v). Instead, leave the transformation in the form u = u(x,y) and v = v(x, y) and
calculate the Jacobian of the original transformation
d(u, v)
d(x,y)
du du
dx dy
dy dy
dx dy
du dv du dv
dx dy dy dx'
Once this is known, we can get the PDF of (U, V) by
(7.7.4)
fu,v(u, v) = fXj(x,y)
1
d(u,v) '
d(x,y)
(7.7.5)
library (emdbook) ; library (mvtnorm) # note: the order matters
mu <- c(Q,0); sigma <- diag(2)
f <- function(x ,y) dmvnorm(c(x,y) , mean = mu, sigma = sigma)
curve3d(f (x,y) , from = c(-3,-3), to = c(3,3), theta = -30, phi = 30)
The code above is slightly shorter than that using persp and is easier to understand. One must be careful, however. If
the library calls are swapped then the code will not work because both packages emdbook and mvtnorm have a function
called “dmvnorm”; one must load them to the search path in the correct order or R will use the wrong one (the arguments
are named differently and the underlying algorithms are different).
182
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
Figure 7.6.1: Graph of a bivariate normal PDF
7.7. BIVARIATE TRANSFORMATIONS OF RANDOM VARIABLES
183
In some cases there will be a cancellation and the work will be a lot shorter. Of course, it is not
always true that
d{x,y) = J_
d (u, v) a(u-v'> ’
d(x,y)
(7.7.6)
but for the well-behaved examples that we will see in this book it works just fine. . . do you see the
connection between Equations 7.7.6 and 6.4.5?
Example 7.29. Let ( X , Y ) ~ mvnorm(mean = 02xi , sigma = I2X2) and consider the transformation
U = 37 + 47,
V = 57 + 67.
We can solve the system of equations to find the inverse transformations; they are
X = - 3U + 2V,
in which case the Jacobian of the inverse transformation is
As (x,y) traverses K2, so too does (m, v). Since the joint PDF of ( X , 7) is
fxj(x, y) = exp |-i (x2 + y2)j , (x,y) e R2,
we get that the joint PDF of (U, V) is
1 1
fu,v(»,v) = ^expl--
(-3m + 2v)2 +
5m - 3v
Remark 7.30. It may not be obvious, but Equation 7.7.7 is the PDF
a more general result see Theorem 7.34.
(m, v) e R2. (7.7.7)
of a mvnorm distribution. For
7.7.1 How to do it with R
It is possible to do the computations above in R with the Ryacas package. The package is an inter-
face to the open-source computer algebra system, “Yacas”. The user installs Yacas, then employs
Ryacas to submit commands to Yacas, after which the output is displayed in the R console.
There are not yet any examples of Yacas in this book, but there are online materials to help the
interested reader: see http : //code . google . com/p/ryacas/ to get started.
184
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
7.8 Remarks for the Multivariate Case
There is nothing spooky about n > 3 random variables. We just have a whole bunch of them: A),
Aj,. . . , X„ , which we can shorten to X = (Aj , X2, . . . , X„ )T to make the formulas prettier (now may
be a good time to check out Appendix E.5). For X supported on the set S the joint PDF /x (if it
exists) satisfies
/x(x)> 0, for x e A x, (7.8.1)
and
/x(x) dvid.r2 • • ■ dx„ = 1 , (7.8.2)
or even shorter: f /x(x) dx = 1 . The joint CDF F\ is debited by
T’x(x) = P(Xi < *i, X2 < x2 < xn), (7.8.3)
for xel". The expectation of a function g(X) is debned just as we would imagine:
Eg(X) = J g(x)fx(x)dx. (7.8.4)
provided the integral exists and is bnite. And the moment generating function in the multivariate
case is debned by
Mx(t) = IE exp {tTX} , (7.8.5)
whenever the integral exists and is bnite for all t in a neighborhood of 0nxi (note that tTX is short-
hand for 1\ X | + jAj + • ■ ■ + t„X„). The only difference in any of the above for the discrete case is
that integrals are replaced by sums.
Marginal distributions are obtained by integrating out remaining variables from the joint distri-
bution. And even if we are given all of the univariate marginals it is not enough to determine the
joint distribution uniquely.
We say that A). Aj, .... X„ are mutually independent if their joint PDF factors into the product
of the marginals
/x(x) = fx,(xi) fx2(x2) ■ ■ ■ /x,,0„), (7.8.6)
for every x in their joint support Sx, and we say that A), Aj, Xn are exchangeable if their joint
PDF (or CDF) is a symmetric function of its n arguments, that is, if
/x(x*) = /x(x), (7.8.7)
for any reordering x* of the elements of x = (xi , x2, . . . , xn) in the joint support.
Proposition 7.31. Let Aj, X2, X„ be independent with respective population means p [ , p2, . . . ,
p„ and standard deviations o~i, cr2, . . . , <x„. For given constants a\, a2, . . . ,an define Y = YIl=\ aiXi-
Then the mean and standard deviation ofY are given by the formulas
n
VY = ^ CliHi, cry
1=1
n
z
i~ 1
2 2.
a : CT:
1/2
(7.8.8)
7.8. REMARKS FOR THE MULTIVARIATE CASE
185
Proof. The mean is easy:
n \ n
ET = E
y ^ ciiXi — y ^ cii e X[ — y ^ ciini.
i=l / i= 1 i= 1
The variance is not too difficult to compute either. As an intermediate step, we calculate E Y 2 .
n \2
Er = E
n— 1 n
2 a* = E 2 + 2 £ Z aiaJXiXJ
i= 1 j=i+ 1
V i=l /
V *=l
Using linearity of expectation the E distributes through the sums. Now EXr = cr2 + p2 and
E XjXj = EX, E Xj = Hifj.j when i + j because of independence. Thus
n n—ln
E Y2 = J] +^ + 2YjX WjMi
i= 1 i=l 7=1+1
Z
2 2 ,
ai<?i +
ft- 1 «
didjllilUj
i=i j=i+i
V i=l
To complete the proof, note that the expression in the parentheses is exactly (E Y)2, and recall the
identity cr~ = E Y2 - (E Y)2. □
There is a corresponding statement of Fact 7.16 for the multivariate case. The proof is also
omitted here.
Fact 7.32. //'X ond Y dre mutually independent random vectors, then u(X) and v(Y) are indepen-
dent for any functions u and v.
Bruno de Finetti was a strong proponent of the subjective approach to probability. He proved
an important theorem in 1931 which illuminates the link between exchangeable random variables
and independent random variables. Here it is in one of its simplest forms.
Theorem 7.33. De Finetti’s Theorem. Let X\, X2, ... be a sequence p/binom(size = 1, prob =
p) random variables such that (X\ , . . . , Xf) are exchangeable for every k. Then there exists a
random variable 0 with support [0, 1] and PDF f@(8) such that
P(X, =X[,...,xk = Xk) = f ez*‘(i-6)k-zx‘M0)de, (7.8.9)
Jo
for all Xj = 0, 1, i = 1, 2, . . . , k.
To get a handle on the intuitive content de Finetti’s theorem, imagine that we have a bunch
of coins in our pocket with each having its own unique value of 8 = P(Heads). We reach into
our pocket and select a coin at random according to some probability - say, f&{8). We take the
randomly selected coin and flip it k times.
Think carefully: the conditional probability of observing a sequence X\ = x\, . . . , X^ = Xk,
given a specific coin 8 would just be ^-^(l - 8)k~^x‘, because the coin flips are an independent
186
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
sequence of Bernoulli trials. But the coin is random, so the Theorem of Total Probability says we
can get the unconditional probability IP (A" | = x\, . . . , = xf) by adding up terms that look like
02*'(1 - Of-Z* f@(0), (7.8.10)
where we sum over all possible coins. The right-hand side of Equation 7.8.9 is a sophisticated way
to denote this process.
Of course, the integral’s value does not change if we jumble the x,’s, so (A) , XU) are clearly
exchangeable. The power of de Finetti’s Theorem is that every infinite binary exchangeable se-
quence can be written in the above form.
The connection to subjective probability: our prior information about 6 corresponds to /©(#)
and the likelihood of the sequence X\ = jti, . . . , X* = x* (conditional on 8) corresponds to -
8)k~ 1* Compare Equation 7.8.9 to Section 4.8 and Section 7.3.
The multivariate normal distribution immediately generalizes from the bivariate case. If the
matrix X is nonsingular then the joint PDF of X ~ mvnorm(mean = p, sigma = X) is
/x(x) =
(2tt)"/2 |X[1/2
exp(--(x- p)T X
'* (x - p) j ,
(7.8.11)
and the MGF is
Mx(t) = exp { pTt + ^tTXt
(7.8.12)
We will need the following in Chapter 12.
Theorem 7.34. //X ~ mvnorm(mean = p, sigma = X) and A is any matrix, then the random
vector Y = AX is distributed
Y ~ mvnorm(mean = Ap, sigma = AXAT). (7.8.13)
Proof. Look at the MGF of Y :
MY(t) = IE exp jtT(AX)} ,
= IE exp {(ATt)Txj ,
- '4I(AT,,+i<AT,)W,)
= exp|(Ap)Tt+itT(AXAT)t|,
and the last expression is the MGF of an mvnorm(mean = Ap, sigma = AXAT) distribution. □
7.9 The Multinomial Distribution
We sample n times, with replacement, from an urn that contains balls of k different types. Let A)
denote the number of balls in our sample of type 1, let X2 denote the number of balls of type 2, . . .
, and let X \ denote the number of balls of type k. Suppose the urn has proportion p\ of balls of type
7. 9. THE MULTINOMIAL DISTRIB UTION
187
1, proportion pi of balls of type 2, . . . , and proportion pk of balls of type k. Then the joint PMF of
(Xi,...,X*)is
/x V. (v, xk) = ( " ) p?p?-p?, (7.9.1)
V*1 X2 ••• Xkj
for (xi , . . . ,Xk) in the joint support S xt,...xK- We write
(Xi, . . ,,Xk) ~ multinom(size = n, prob = pkxi). (7.9.2)
Several comments are in order. First, the joint support set Sxu..jcx contains all nonnegative
integer k-tuples (x\, . . . , Xk) such that X\ + x2 + • • • + Xk = n. A support set like this is called a
simplex. Second, the proportions p\ , pi, Pk satisfy p, > 0 for all i and p\ + pi + ■■■+ pi, = 1 •
Finally, the symbol
n
x\ x2 ■■■ xk
is called a multinomial coefficient which generalizes the notion of a binomial coefficient we saw in
Equation 4.5.1.
The form and notation we have just described matches the R usage but is not standard among
other texts. Most other books use the above for a k - 1 dimension multinomial distribution, because
the linear constraint x\+ x2 + ■ ■ ■ + xk — n means that once the values of Xi, Xj, ■ . . , X*_i are known
the final value Xk is determined, not random. Another term used for this is a singular distribution.
For the most part we will ignore these difficulties, but the careful reader should keep them
in mind. There is not much of a difference in practice, except that below we will use a two-
dimensional support set for a three-dimension multinomial distribution. See Figure 7.9.1.
When k — 2, we have x\ = x and X2 = n-x, we have p\ = p and pi = 1 —p, and the multinomial
coefficient is literally a binomial coefficient. In the previous notation we have thus shown that the
multinomfsize = n, prob = P2xi) distribution is the same as a binomfsize = n, prob = p)
distribution.
Example 7.35. Dinner with Barack Obama. During the 2008 U.S. presidential primary, Barack
Obama offered to have dinner with three randomly selected monetary contributors to his campaign.
Imagine the thousands of people in the contributor database. For the sake of argument. Suppose that
the database was approximately representative of the U.S. population as a whole. Suppose Barack
Obama wants to have dinner http://pewresearch.org/pubs/773/fewer-voters-identify-
36 democrat, 27 republican , 37 independent.
Remark 7.36. Here are some facts about the multinomial distribution.
1. The expected value of (X1; X2, . . . , Xk) is npkx\.
2. The variance-covariance matrix 2 is symmetric with diagonal entries cr ? = np,(] - /?,),
i = 1, 2, . . . , k and off-diagonal entries Cov(X,, Xj) = - npiPj , for i 4- j. The correlation
between X, and Xj is therefore Corr(X,-, Xj) = - yjpiPj/ (1 - pi){ 1 - pj).
3. The marginal distribution of (X1; X2, ..., Xk-j) is multinom(size = n, prob = P(t-i)xi)
with
P(*-i)xi = (P 1. Pi, ■■■, Pk—i, Pk-i + Pk) ,
and in particular, X,- ~ binom(size = n, prob = pi).
x1!x2! ■ ■■ xk !
(7.9.3)
(7.9.4)
188
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
7.9.1 How to do it with R
There is support for the multinomial distribution in base R, namely in the stats package. The
dmultinom function represents the PMF and the rmultinom function generates random variates.
> library (combinat)
> tmp <- t(xsimplex(3 , 6 ))
> p <- applyCtmp , MARGIN = 1, FUN = dmultinom, prob = c(36,
+ 27, 37))
> library (prob)
> S <- probspace(tmp, probs = p)
> ProbTable <- xtabs(probs ~ XI + X2 , data = S)
> round(ProbTable , 3)
X2
«
1
2
3
4
5
6
0
8.
.883
0.
.011
0.
.020
0,
.020
0,
.011
0.
.003
0.
.00®
1
0.
.015
0.
.055
0.
.08®
0.
.058
0,
.021
0.
.003
0.
.000
2
0.
.036
0.
.106
0.
.116
0.
.057
0,
.010
0.
.000
0.
.00®
B
®.
.047
0.
.103
0.
.076
0.
.018
0,
.000
0.
.000
0.
.00®
4
®.
.034
0.
.050
0.
.018
0.
.000
0,
.000
0.
.000
0.
.00®
5
®.
.013
0.
.010
0.
.00®
0.
.000
0,
.000
0.
.000
0.
.00®
6
®.
.002
0.
.000
0.
.00®
0.
.000
0,
.000
0.
.000
0.
.00®
Do some examples of rmultinom.
Here is another way to do it4.
4Another way to do the plot is with the scatterplot3d function in the scatterplot3d package [61]. It looks like
this:
library(scatterplot3d)
X <- t(as.matrix(expand.grid(®:6, 8:6)))
X <- X[ , colSums(X) <= 6] ; X <- rbind(X, 6 - colSums(X))
Z <- round(apply(X, 2, function(x) dmultinomfx, prob = 1:3)), 3)
A <- data.frame(x = X[l, ], y = X[2, ], probability = Z)
scatterplot3d(A, type = “h” , lwd = 3, box = FALSE)
The scatterplot3d graph looks better in this example, but the code is clearly more difficult to understand. And with
cloud one can easily do conditional plots of the form cloud(z ~x + y | f) , where f is a factor.
7. 9. THE MULTINOMIAL DISTRIB UTION
189
Figure 7.9.1: Plot of a multinomial PMF
190
CHAPTER 7. MULTIVARIATE DISTRIBUTIONS
Chapter Exercises
Exercise 7.1. Prove that Covt X, Y) = E (XY) - (EX)(E Y).
Exercise 7.2. Suppose X ~ chisq(df = p\) and Y ~ chisq(df = po) are independent. Find the
distribution of X + Y (you may want to refer to Equation 6.5.6).
Exercise 7.3. Show that when X and Y are independent the MGF of X - Y is Mx(t)MY(-t). Use
this to find the distribution of X - Y when X ~ norm(mean = p\, sd = cri) and Y ~ norm(mean =
P2, sd = cr2) are independent.
Chapter 8
Sampling Distributions
This is an important chapter; it is the bridge from probability and descriptive statistics that we
studied in Chapters 3 through 7 to inferential statistics which forms the latter part of this book.
Here is the link: we are presented with a population about which we would like to learn. And
while it would be desirable to examine every single member of the population, we find that it is
either impossible or infeasible to for us to do so, thus, we resort to collecting a sample instead. We
do not lose heart. Our method will suffice, provided the sample is representative of the population.
A good way to achieve this is to sample randomly from the population.
Supposing for the sake of argument that we have collected a random sample, the next task
is to make some sense out of the data because the complete list of sample information is usually
cumbersome, unwieldy. We summarize the data set with a descriptive statistic, a quantity calculated
from the data (we saw many examples of these in Chapter 3). But our sample was random. . .
therefore, it stands to reason that our statistic will be random, too. How is the statistic distributed?
The probability distribution associated with the population (from which we sample) is called
the population distribution, and the probability distribution associated with our statistic is called its
sampling distribution; clearly, the two are interrelated. To learn about the population distribution,
it is imperative to know everything we can about the sampling distribution. Such is the goal of this
chapter.
We begin by introducing the notion of simple random samples and cataloguing some of their
more convenient mathematical properties. Next we focus on what happens in the special case of
sampling from the normal distribution (which, again, has several convenient mathematical prop-
erties), and in particular, we meet the sampling distribution of X and S2. Then we explore what
happens to A’s sampling distribution when the population is not normal and prove one of the most
remarkable theorems in statistics, the Central Limit Theorem (CLT).
With the CLT in hand, we then investigate the sampling distributions of several other popular
statistics, taking full advantage of those with a tractable form. We finish the chapter with an ex-
ploration of statistics whose sampling distributions are not quite so tractable, and to accomplish
this goal we will use simulation methods that are grounded in all of our work in the previous four
chapters.
What do I want them to know?
191
192
CHAPTER 8. SAMPLING DISTRIBUTIONS
• the notion of population versus simple random sample, parameter versus statistic, and popu-
lation distribution versus sampling distribution
• the classical sampling distributions of the standard one and two sample statistics
• how to generate a simulated sampling distribution when the statistic is crazy
• the Central Limit Theorem, period.
• some basic concepts related to sampling distribution utility, such as bias and variance
8.1 Simple Random Samples
8.1.1 Simple Random Samples
Definition 8.1. If X\, Xn, . . . , X„ are independent with X, ~ f for i = 1,2,..., n, then we say
that X\ , Xn, ..., Xn are independent and identically distributed (i.i.d.) from the population / or
alternatively we say that X\, Xn, . . . , Xn are a simple random sample of size n, denoted SRS (n),
from the population /.
Proposition 8.2. Let X\, X2, X„ be a SRS (n) from a population distribution with mean p and
finite standard deviation cr. Then the mean and standard deviation ofX are given by the formulas
Px~P and = cr/ yfn.
Proof Plug in a\ = 02 = •••= an — \/n in Proposition 7.31. □
The next fact will be useful to us when it comes time to prove the Central Limit Theorem in
Section 8.3.
Proposition 8.3. Let X\, Xn, X„ be a SRS (n) from a population distribution with MGF M(t).
Then the MGF ofX is given by
And because X\, Xn, . . . , Xn are independent. Proposition 7.13 allows us to distribute the expecta-
tion among each term in the product, which is
(8.1.1)
Proof Go from the definition:
Mx(t) = IE elX,
— jg Qt(.Xi+—+X„)/n
— jg QtXi/n^tX2/n m # e gtXn/n
IE e'x' /n IE e'X2/" • ■ • IE e,tXJn
The last step is to recognize that each term in last product above is exactly M(t/n).
□
8.2. SAMPLING FROM A NORMAL DISTRIB UTION
193
8.2 Sampling from a Normal Distribution
8.2.1 The Distribution of the Sample Mean
Proposition 8.4. Let X i, X2, ..., X„ be a SRS(n)from a norm(mean = p, sd = cr) distribution.
Then the sample mean X has a norm(mean - p. sd - cr / -fn) sampling distribution.
Proof. The mean and standard deviation of X follow directly from Proposition 8.2. To address the
shape, first remember from Section 6.3 that the norm(mean = p. sd = cr) MGF is of the form
M(t ) = exp J pt + crV/2} .
Now use Proposition 8.3 to find
Mj(t)
[exp [p{t/n) + cr20/n)2/2}]” ,
exp [ n ■ [p(t/n) + cr2(f/n)2/2]} ,
exp {y ut + (cr/ \fn)2t2 / 2} ,
and we recognize this last quantity as the MGF of a norm(mean = p, sd = cr/ yfn) distribution. □
8.2.2 The Distribution of the Sample Variance
Theorem 8.5. Let X\, X2, ..., Xn be a S RS ( n)from a norm(mean = p, sd - cr) distribution, and
let
n j n
X = V Xi and S2 = V (Xi - X)2 . (8.2.1)
X—i i n — 1 X—t i
i=i i=i
Then
1. X and S2 are independent, and
2. The rescaled sample variance
(n- 1) 2_ XU*- X?
cr2 cr2
has a chisqldf = n — 1) sampling distribution.
(8.2.2)
Proof. The proof is beyond the scope of the present book, but the theorem is simply too important
to be omitted. The interested reader could consult Casella and Berger [13], or Hogg et al [43]. □
8.2.3 The Distribution of Student’s T Statistic
Proposition 8.6. Let Xi, X2, ..., X„ be a SRS(n)from a norm(mean = p, sd = cr) distribution.
Then the quantity
T =
X-p
S / sfn
(8.2.3)
has a t(df — n — 1) sampling distribution.
194
CHAPTER 8. SAMPLING DISTRIBUTIONS
Proof. Divide the numerator and denominator by cr and rewrite
x-n
w v«
S /cr
x-n
<r/ y/n
fWfC- 1)'
Now let
so that
and V =
(n - 1)5 2
cr-
T =
Z
~Wr’
(8.2.4)
where r = n — 1 .
We know from Section 8.2.1 thatZ ~ norm(mean = 0, sd = 1) and we know from Section 8.2.2
that V ~ chisq(df = n— 1). Further, since we are sampling from a normal distribution. Theorem 8.5
gives that X and S 2 are independent and by Fact 7.16 so are Z and V. In summary, the distribution
of T is the same as the distribution of the quantity Z/ sJV/r, where Z ~ norm(mean = 0, sd = 1)
and V ~ chisq(df = r) are independent. This is in fact the definition of Student’s t distribution. □
This distribution was first published by W. S. Gosset (1900) under the pseudonym Student, and
the distribution has consequently come to be known as Student’s t distribution. The PDF of T can
be derived explicitly using the techniques of Section 6.4; it takes the form
fx(x) =
T[(r + l)/2]
sfrn T(r/2)
|-(r+l)/2
— OO < x < oo
(8.2.5)
Any random variable X with the preceding PDF is said to have Student’s t distribution with r
degrees of freedom, and we write X ~ t(df = r). The shape of the PDF is similar to the normal,
but the tails are considerably heavier. See Figure 8.2.1. As with the normal distribution, there are
four functions in R associated with the t distribution, namely dt, pt, qt, and rt, which compute
the PDF, CDF, quantile function, and generate random variates, respectively.
The code to produce Figure 8.2.1 is
> curve(dt(x, df = 30), from = -3, to
> ind <- c(l, 2, 3, 5, 10)
> for (i in ind) curve(dt(x, df = i) ,
= 3, lwd = 3, ylab = "y")
-3, 3, add = TRUE)
Similar to that done for the normal we may define t,,(df — n — 1) as the number on the x-axis
such that there is exactly a area under the t( df = n - 1) curve to its right.
Example 8.7. Find t0.oi(df = 23) with the quantile function.
> qt(0.01, df = 23, lower . tail = FALSE)
[1] 2.499867
Remark 8.8. There are a few things to note about the t(df = r) distribution.
8.2. SAMPLING FROM A NORMAL DISTRIB UTION
195
Figure 8.2.1: Student’s t distribution for various degrees of freedom
196
CHAPTER 8. SAMPLING DISTRIBUTIONS
1. The t(df = 1) distribution is the same as the cauchydocation = 0, scale = 1) distribu-
tion. The Cauchy distribution is rather pathological and is a counterexample to many famous
results.
2. The standard deviation of t(df = r) is undefined (that is, infinite) unless r > 2. When r is
more than 2, the standard deviation is always bigger than one, but decreases to 1 as r — » oo.
3. As r — » oo, the t(df = r) distribution approaches the norm(mean = 0, sd = 1) distribution.
8.3 The Central Limit Theorem
In this section we study the distribution of the sample mean when the underlying distribution is not
normal. We saw in Section 8.2 that when Xi, Xo, . . . , Xn is a SRS (n) from a norm(mean = p, sd =
cr) distribution then X ~ norm(mean - p, sd = cr/ sfn). In other words, we may say (owing to Fact
6.24) when the underlying population is normal that the sampling distribution of Z defined by
Z =
X-p
< t / sjn
(8.3.1)
is norm(mean = 0, sd = 1).
However, there are many populations that are not normal. . . and the statistician often finds
herself sampling from such populations. What can be said in this case? The surprising answer is
contained in the following theorem.
Theorem 8.9. The Central Limit Theorem. Let X\, Xj, Xn be a SRS ( n ) from a population
distribution with mean p and finite standard deviation cr. Then the sampling distribution of
approaches a norm(mean = 0, sd = 1) distribution as n — > oo.
Remark 8.10. We suppose that Xi, X2, .... X„ are i.i.d., and we learned in Section 8.1.1 that X has
mean p and standard deviation cr/ yfn, so we already knew that Z has mean 0 and standard deviation
1. The beauty of the CLT is that it addresses the shape of Z’s distribution when the sample size is
large.
Remark 8.11. Notice that the shape of the underlying population’s distribution is not mentioned in
Theorem 8.9; indeed, the result is true for any population that is well-behaved enough to have a
finite standard deviation. In particular, if the population is normally distributed then we know from
Section 8.2. 1 that the distribution of X (and Z by extension) is exactly normal, for every n.
Remark 8.12. How large is “sufficiently large”? It is here that the shape of the underlying popula-
tion distribution plays a role. For populations with distributions that are approximately symmetric
and mound-shaped, the samples may need to be only of size four or five, while for highly skewed or
heavy-tailed populations the samples may need to be much larger for the distribution of the sample
means to begin to show a bell-shape. Regardless, for a given population distribution (with finite
standard deviation) the approximation tends to be better for larger sample sizes.
8.4. SAMPLING DISTRIBUTIONS OF TWO-SAMPLE STATISTICS
197
8.3.1 How to do it with R
The TeachingDemos package [79] has clt . examp and the distrTeach [74] package has illustrateCLT.
Try the following at the command line (output omitted):
> library (TeachingDemos)
> exampl e(clt. examp)
> library(distrTeach)
> example (illustrateCLT)
The IPSUR package has the functions cltl, clt2, and cltB (see Exercise 8.2 at the end of
this chapter). Its purpose is to investigate what happens to the sampling distribution of X when the
population distribution is mound shaped, finite support, and skewed, namely t(df = 3), unif(a =
0, b = 10) and gamma( shape = 1.21, scale = 1/2.37), respectively.
For example, when the command clt 1 Q is issued a plot window opens to show a graph of the
PDF of a t( df = 3) distribution. On the display are shown numerical values of the population mean
and variance. While the students examine the graph the computer is simulating random samples of
size sample . size = 2 from the population = "rt" distribution a total of N . iter = 180080
times, and sample means are calculated of each sample. Next follows a histogram of the simulated
sample means, which closely approximates the sampling distribution of X, see Section 8.5. Also
show are the sample mean and sample variance of all of the simulated As. As a final step, when the
student clicks the second plot, a normal curve with the same mean and variance as the simulated
As is superimposed over the histogram. Students should compare the population theoretical mean
and variance to the simulated mean and variance of the sampling distribution. They should also
compare the shape of the simulated sampling distribution to the shape of the normal distribution.
The three separate cltl, clt2, and clt3 functions were written so that students could compare
what happens overall when the shape of the population distribution changes. It would be possible
to combine all three into one big function, clt which covers all three cases (and more).
8.4 Sampling Distributions of Two-Sample Statistics
There are often two populations under consideration, and it sometimes of interest to compare prop-
erties between groups. To do so we take independent samples from each population and calculate
respective sample statistics for comparison. In some simple cases the sampling distribution of the
comparison is known and easy to derive; such cases are the subject of the present section.
8.4.1 Difference of Independent Sample Means
Proposition 8.13. Let Ai, A2, .... A„, be an S RS (n \ ) from a norm(mean = px, sd = <:rx) dis-
tribution and let Y\, Y2, ..., Y„2 be an SRS(n2)from a norm(mean = / uy, sd = cry) distribution.
Suppose that X\, X2, . . . , A„, and Y\, Y2, . . . , Yni are independent samples. Then the quantity
and
X - Y - ( px ~ Py)
(8.4.1)
198
CHAPTER 8. SAMPLING DISTRIBUTIONS
has a norm(mean = 0, sd = 1) sampling distribution. Equivalently, X — Y has a norm(mean =
px ~ A 1y, sd = -J «i + ?l) tsatnpling distribution.
Proof. We know that X is norm(mean = px, sd = crx/ x/hi) and we also know that Y is normfmean =
Py, sd = cry I xfnf). And since the samples X\, X2, . . . , X,h and Y 1, V2, . . . , Y„, are independent,
so too are X and Y. The distribution of their difference is thus normal as well, and the mean and
standard deviation are given by Proposition 7.20. □
Remark 8.14. Even if the distribution of one or both of the samples is not normal, the quantity in
Equation 8.4.1 will be approximately normal provided both sample sizes are large.
Remark 8. 15. For the special case of px = Py we have shown that
, (8.4.2)
xjo'x/n 1 +cr2Y/n2
has a normfmean = 0, sd = 1) sampling distribution, or in other words, X-Y has a normimean =
0, sd = Jcr^/ni +<Ty/n 2) sampling distribution. This will be important when it comes time to do
hypothesis tests; see Section 9.3.
8.4.2 Difference of Independent Sample Proportions
Proposition 8.16. Let Xi, X2, ..., X„t be an S RS (nf) from a binomfsize = 1, prob = pf) dis-
tribution and let Y\, Y2, . . . , Y„2 be an SRS (n2) from a binomfsize = 1, prob = p2) distribution.
Suppose that X\, X2, . . . , X,n and Y\, Y2, . . . , T„, are independent samples. Define
2 «i 1 "2
Pi = — YX- and p2 = — Y Yj.
Then the sampling distribution of
Pi ~P2 ~ (P\ ~ Pi)
V
"1
+
Pl(l-Pl)
>12
(8.4.3)
(8.4.4)
approaches a norm(mean = 0, sd = 1) distribution as both n\, n2 — > 00. In other words, the
sampling distribution of p\ - p2 is approximately
norm
mean - p\ - p2, sd =
Ip i(l ~ Pi) , PiO - Pi)
n 1
n2
provided both n \ and n2 are sufficiently large.
(8.4.5)
Proof. We know that p\ is approximately normal for n \ sufficiently large by the CLT, and we know
that p2 is approximately normal for n2 sufficiently large, also by the CLT. Further, p\ and p2 are
independent since they are derived from independent samples. And a difference of independent
8.4. SAMPLING DISTRIBUTIONS OF TWO-SAMPLE STATISTICS
199
(approximately) normal distributions is (approximately) normal, by Exercise 7.3 1 . The expressions
for the mean and standard deviation follow immediately from Proposition 7.20 combined with the
formulas for the binom(size = 1, prob = p) distribution from Chapter 5. □
8.4.3 Ratio of Independent Sample Variances
Proposition 8.18. Let Xi, X2, .... Xm be an S RS («i) from a norm(mean = px. sd = <:rx ) dis-
tribution and let Y\, 13. •••. Yni be an SRS {nf) from a norm(mean = py, sd = cry) distribution.
Suppose that X\, X 2, .... Xm and Y\, Y2, .... Y„2 are independent samples. Then the ratio
F =
°T-S|
tr\S\
(8.4.6)
has an f(dfl = n \ - 1, df2 = n2 - 1) sampling distribution.
Proof. We know from Theorem 8.5 that (n\ - 1 )S1 2x/cr2x is distributed chisq(df = n\ - 1) and
(n2 - 1 )SY/crY is distributed chisq(df = n2 ~ 1)- Now write
_ cr2Ysl _ (m - 1 )S2yl(>n - 1) Ua*
cr2xS2. (n2- ] )S2Yj (n2 - 1) V o\'
by multiplying and dividing the numerator with n\— 1 and doing likewise for the denominator with
«2 - 1. Now we may regroup the terms into
F =
(m-l)Sf
°i \
l(n 1 - 1)
°Y 1
1 (n2 - 1)
and we recognize F to be the ratio of independent chisq distributions, each divided by its respective
numerator df = n\ - 1 and denominator df = n\ - 1 degrees of freedom. This is, indeed, the
definition of Snedecor’s F distribution. □
Remark 8.19. For the special case of crx = cry we have shown that
S 2
F=- § (8.4.7)
SY
has an f(df 1 = n \ - 1, df2 = ni_ - 1 ) sampling distribution. This will be important in Chapters 9
onward.
1
Remark 8.17. This does not explicitly follow, because of our cavalier use of “approximately” in too many places. To be
more thorough, however, would require more concepts than we can afford at the moment. The interested reader may consult
a more advanced text, specifically the topic of weak convergence, that is, convergence in distribution.
200
CHAPTER 8. SAMPLING DISTRIBUTIONS
8.5 Simulated Sampling Distributions
Some comparisons are meaningful, but their sampling distribution is not quite so tidy to describe
analytically. What do we do then?
As it turns out, we do not need to know the exact analytical form of the sampling distribution;
sometimes it is enough to approximate it with a simulated distribution. In this section we will show
you how. Note that R is particularly well suited to compute simulated sampling distributions, much
more so than, say, SPSS or SAS.
8.5.1 The Interquartile Range
> iqrs <- replicate (ISO, IQR(rnorm(lSQ)))
We can look at the mean of the simulated values
> mean(iqrs) # close to 1
[1] 1.BB9361
and we can see the standard deviation
> sd(iqrs)
[1] 0.1613881
Now let’s take a look at a plot of the simulated values
8.5.2 The Median Absolute Deviation
> mads <- replicateCUSS, mad(rnorm(H9IS)))
We can look at the mean of the simulated values
> mean (mads) # close to 1.349
[1] 1.803938
and we can see the standard deviation
> sd(mads)
[1] 8.1213811
Now let’s take a look at a plot of the simulated values
Frequency
8.5. SIMULATED SAMPLING DISTRIBUTIONS
201
Histogram of iqrs
1.0 1.2 1.4 1.6 1.8
iqrs
Figure 8.5.1: Plot of simulated IQRs
202
CHAPTER 8. SAMPLING DISTRIBUTIONS
Histogram of mads
>
o
c=
CD
=3
cr
CD
0.8 0.9 1.0 1.1 1.2 1.3
mads
Figure 8.5.2: Plot of simulated MADs
8.5. SIMULATED SAMPLING DISTRIBUTIONS
203
Chapter Exercises
Exercise 8.1. Suppose that we observe a random sample X\, Xo, . . . , Xn of size SRS (n — 26 ) from
a normfmean =19) distribution.
1 . What is the mean of XI
2. What is the standard deviation of XI
3. What is the distribution of XI (approximately)
4. Find P(n < X < b)
5. Find P(X > c).
Exercise 8.2. In this exercise we will investigate how the shape of the population distribution
affects the time until the distribution of X is acceptably normal.
Answer the questions and write a report about what you have learned. Use plots and histograms to
support your conclusions. See Appendix F for instructions about writing reports with R. For these
problems, the discussion/interpretation parts are the most important, so be sure to ANSWER THE
WHOLE QUESTION.
The Central Limit Theorem
For Questions 1-3, we assume that we have observed random variables X\, 2Q, . . . ,X„ that are
an SRS(n ) from a given population (depending on the problem) and we want to investigate the
distribution of X as the sample size n increases.
1 . The population of interest in this problem has a Student’s t distribution with r - 3 degrees of
freedom. We begin our investigation with a sample size of n = 2. Open an R session, make
sure to type library(IPSUR) and then follow that with cltl().
(a) Look closely and thoughtfully at the first graph. How would you describe the popula-
tion distribution? Think back to the different properties of distributions in Chapter 3.
Is the graph symmetric? Skewed? Does it have heavy tails or thin tails? What else can
you say?
(b) What is the population mean /j and the population variance cr2? (Read these from the
first graph.)
(c) The second graph shows (after a few seconds) a relative frequency histogram which
closely approximates the distribution of X. Record the values of mean(xbar) and
var(xbar), where xbar denotes the vector that contains the simulated sample means.
Use the answers from part (b) to calculate what these estimates should be, based on
what you know about the theoretical mean and variance of X. How well do your an-
swers to parts (b) and (c) agree?
(d) Click on the histogram to superimpose a red normal curve, which is the theoretical
limit of the distribution of X as n — > oo. How well do the histogram and the normal
curve match? Describe the differences between the two distributions. When judging
204
CHAPTER 8. SAMPLING DISTRIBUTIONS
between the two, do not worry so much about the scale (the graphs are being rescaled
automatically, anyway). Rather, look at the peak: does the histogram poke through the
top of the normal curve? How about on the sides: are there patches of white space
between the histogram and line on either side (or both)? How do the curvature of the
histogram and the line compare? Check down by the tails: does the red line drop off
visibly below the level of the histogram, or do they taper off at the same height?
(e) We can increase our sample size from 2 to 1 1 with the command cltl (sample . size
= 11). Return to the command prompt to do this. Answer parts (b) and (c) for this new
sample size.
(f) Go back to cltl and increase the sample . size from 11 to 31. Answer parts (b) and
(c) for this new sample size.
(g) Comment on whether it appears that the histogram and the red curve are “noticeably
different” or whether they are “essentially the same” for the largest sample size n = 31.
If they are still “noticeably different” at n = 31, how large does n need to be until they
are “essentially the same”? (Experiment with different values of n).
2. Repeat Question 1 for the function clt2. In this problem, the population of interest has a
uniffmin = 0, max = 10) distribution.
3. Repeat Question 1 for the function clt3. In this problem, the population of interest has a
gamma( shape = 1.21, rate = 1/2.37) distribution.
4. Summarize what you have learned. In your own words, what is the general trend that is
being displayed in these histograms, as the sample size n increases from 2 to 11, on to 31,
and onward?
5. How would you describe the relationship between the shape of the population distribution
and the speed at which A’s distribution converges to normal? In particular, consider a pop-
ulation which is highly skewed. Will we need a relatively large sample size or a relatively
small sample size in order for A’s distribution to be approximately bell shaped?
Exercise 8.3. Let X\,. . . , X22 be a random sample from a normimean = 37, sd = 45) distribution,
and let X be the sample mean of these n = 25 observations. Find the following probabilities.
1. How is X distributed?
norm(mean = 37, sd = 45/ V25)
2. Find P(X> 43.1).
> pnorm(43 . 1 , mean = 37, sd = 9, lower. tail = FALSE)
[1] 8.2489563
Chapter 9
Estimation
We will discuss two branches of estimation procedures: point estimation and interval estimation.
We briefly discuss point estimation first and then spend the rest of the chapter on interval estimation.
We find an estimator with the methods of Section 9.1. We make some assumptions about the
underlying population distribution and use what we know from Chapter 8 about sampling distribu-
tions both to study how the estimator will perform, and to find intervals of confidence for underlying
parameters associated with the population distribution. Once we have confidence intervals we can
do inference in the form of hypothesis tests in the next chapter.
What do I want them to know?
• how to look at a problem, identify a reasonable model, and estimate a parameter associated
with the model
• about maximum likelihood, and in particular, how to
o eyeball a likelihood to get a maximum
o use calculus to find an MLE for one -parameter families
• about properties of the estimators they find, such as bias, minimum variance, MSE
• point versus interval estimation, and how to find and interpret confidence intervals for basic
experimental designs
• the concept of margin of error and its relationship to sample size
9.1 Point Estimation
The following example is how I was introduced to maximum likelihood.
Example 9.1. Suppose we have a small pond in our backyard, and in the pond there live some fish.
We would like to know how many fish live in the pond. How can we estimate this? One procedure
developed by researchers is the capture-recapture method. Here is how it works.
205
206
CHAPTER 9. ESTIMATION
We will fish from the pond and suppose that we capture M = 7 fish. On each caught fish we
attach an unobtrusive tag to the fish’s tail, and release it back into the water.
Next, we wait a few days for the fish to remix and become accustomed to their new tag. Then
we go fishing again. On the second trip some of the fish we catch may be tagged; some may not be.
Let X denote the number of caught fish which are tagged1, and suppose for the sake of argument
that we catch K — 4 fish and we find that 3 of them are tagged.
Now let F denote the (unknown) total number of fish in the pond. We know that F >7, because
we tagged that many on the first trip. In fact, if we let N denote the number of untagged fish in the
pond, then F — M + N . We have sampled K — 4 times, without replacement, from an urn which
has M — 1 white balls and N = F-M black balls, and we have observed x = 3 of them to be white.
What is the probability of this?
Looking back to Section 5.6, we see that the random variable X has a hypertm = M, n =
F - M, k = K) distribution. Therefore, for an observed value X — x the probability would be
JP(X = x) =
("XL")
First we notice that F must be at least 7. Could F be equal to seven? If F = 1 then all of the
fish would have been tagged on the first run, and there would be no untagged fish in the pond, thus,
P(3 successes in 4 trials) = 0.
What about F — 8; what would be the probability of observing X = 3 tagged fish?
P(3 successes in 4 trials) =
35
= — = 0.5.
70
Similarly, if F = 9 then the probability of observing X — 3 tagged fish would be
P(3 successes in 4 trials) =
70
126
0.556.
We can see already that the observed data X — 3 is more likely when F — 9 than it is when F = 8.
And here lies the genius of Sir Ronald Aylmer Fisher: he asks, “What is the value of F which has
the highest likelihood?” In other words, for all of the different possible values of F, which one
makes the above probability the biggest? We can answer this question with a plot of P(A = x)
versus F. See Figure 9.1.1.
Example 9.2. In the last example we were only concerned with how many fish were in the pond,
but now, we will ask a different question. Suppose it is known that there are only two species of
fish in the pond: smallmouth bass ( Micropterus dolomieu) and bluegill (Lepomis macrochirus)',
perhaps we built the pond some years ago and stocked it with only these two species. We would
like to estimate the proportion of fish in the pond which are bass.
Let p = the proportion of bass. Without any other information, it is conceivable for p to be
any value in the interval [0, 1], but for the sake of argument we will suppose that p falls strictly
'it is theoretically possible that we could catch the same tagged fish more than once, which would inflate our count of
tagged fish. To avoid this difficulty, suppose that on the second trip we use a tank on the boat to hold the caught fish until
data collection is completed.
Likelihood
9.1. POINT ESTIMATION
207
number of fish in pond
Figure 9.1.1: Capture-recapture experiment
208
CHAPTER 9. ESTIMATION
between 0 and 1. How can we learn about the true value of pi Go fishing! As before, we will use
catch-and-release, but unlike before, we will not tag the fish. We will simply note the species of
any caught fish before returning it to the pond.
Suppose we catch n fish. Let
(1, if the ith fish is a bass,
0, if the ith fish is a bluegill.
Since we are returning the fish to the pond once caught, we may think of this as a sampling
scheme with replacement where the proportion of bass p does not change. Given that we allow
the fish sufficient time to “mix” once returned, it is not completely unreasonable to model our
fishing experiment as a sequence of Bernoulli trials, so that the X/s would be i.i.d. binomfsize =
1, prob = p). Under those assumptions we would have
PCX) = xi, X2 = x2, Xn = xn) = P(Xi = X|) P(X2 = x2) ■ ■ ■ P(X„ = xn),
= p"{ 1 - pf' pxY\ - pf2 • • • px" ( 1 - pf\
= p2*'(l -pf~^x>.
That is,
PGL = X\ , x2 = X2, Xn = Xn) = pZ*( l - p)n~^x‘.
This last quantity is a function of p, called the likelihood function L(p)\
z<p) = pz*(i-pr2*.
A graph of L for values of £ x, = 3, 4, and 5 when n = 7 is shown in Figure 9.1.2.
> curve(xA5 * (1 - x)A2, from = Q, to = 1, xlab = "p", ylab = "L(p)")
> curve(xA4 * (1 - x)A3, from =6, to = 1 , add = TRUE )
> curve(xA3 * (1 - x)A4, ®, 1, add = TRUE)
We want the value of p which has the highest likelihood, that is, we again wish to maximize
the likelihood. We know from calculus (see Appendix E.2) to differentiate L and set L' — 0 to find
a maximum.
L'(p) = (YJXi)pZxi~1^-p)n~Z
The derivative vanishes (I! = 0) when
{^x^p^-'O ~ p)ll~Lx'
2>d “ P)
Yix‘-pYiXi
+ pLXi -prlx‘-'(-y).
p^(n-YJ^i)^-prZx,-\
{n~YJXI)p,
np ~ p'Zj Xh
P-
This “best” p, the one which maximizes the likelihood, is called the maximum likelihood estimator
(MLE) of p and is denoted p. That is,
P =
l ■A'l
n
= x.
(9.1.1)
9.1. POINT ESTIMATION
209
LO
O
O
O
O
O
O
0.0 0.2 0.4
0.6 0.8
1.0
P
Figure 9.1.2: Assorted likelihood functions for fishing, part two
Three graphs are shown of L when 2 xt equals 3. 4, and 5, respectively, from left to right. We pick an L that
matches the observed data and then maximize Las a function of p. If 2 •*; = 4, then the maximum appears to
occur somewhere around p ~ 0.6.
210
CHAPTER 9. ESTIMATION
parameter space
Figure 9.1.3: Species maximum likelihood
Remark 9.3. Strictly speaking we have only shown that the derivative equals zero at p, so it is
theoretically possible that the critical value p — x is located at a minimum2 instead of a maximum!
We should be thorough and check that L' > 0 when p < x and L' < 0 when p > x. Then by the
First Derivative Test (Theorem E.6) we could be certain that p — x is indeed a maximum likelihood
estimator, and not a minimum likelihood estimator.
The result is shown in Figure 9.1.3.
In general, we have a family of PDFs f(x\6) indexed by a parameter 6 in some parameter space
0. We want to learn about 6. We take a SRS (n):
X\, Xi, . . . ,Xn which are i.i.d. f(x|0).
(9.1.2)
2We can tell from the graph that our value of p is a maximum instead of a minimum so we do not really need to worry
for this example. Other examples are not so easy, however, and we should be careful to be cognizant of this extra step.
9.1. POINT ESTIMATION
211
Definition 9.4. Given the observed data x\,x2, xn, the likelihood function L is defined by
n
L(6) = y\f{Xi\e), 8e@.
i=i
The next step is to maximize L. The method we will use in this book is to find the derivative L'
and solve the equation L’(ff) = 0. Call a solution 9. We will check that L is maximized at 8 using
the First Derivative Test or the Second Derivative Test ( L"(8 ) < (j).
Definition 9.5. A value 8 that maximizes L is called a maximum likelihood estimator (MLE) and
is denoted 8. It is a function of the sample, 9 = 9(X\, X2, . . . , Xn), and is called a point estimator
of 8.
Remark 9.6. Some comments about maximum likelihood estimators:
• Often it is easier to maximize the log-likelihood 1(6) = In L(0) instead of the likelihood L.
Since the logarithmic function y = In x is a monotone transformation, the solutions to both
problems are the same.
• MLEs do not always exist (for instance, sometimes the likelihood has a vertical asymptote),
and even when they do exist, they are not always unique (imagine a function with a bunch of
humps of equal height). For any given problem, there could be zero, one, or any number of
values of 8 for which L(8 ) is a maximum.
• The problems we encounter in this book are all very nice with likelihood functions that
have closed form representations and which are optimized by some calculus acrobatics. In
practice, however, likelihood functions are sometimes nasty in which case we are obliged to
use numerical methods to find maxima (if there are any).
• MLEs are just one of many possible estimators. One of the more popular alternatives are the
method of moments estimators-, see Casella and Berger [13] for more.
Notice, in Example 9.2 we had Xj i.i.d. binom(size = 1, prob = p ), and we saw that the MLE
was p — X . But further
EX
Xi + X2 + ■ ■ ■ + Xn
JE ,
n
-(IEXi + EX2 + ■■■ + EXn),
n
- (np) ,
n
P,
which is exactly the same as the parameter which we estimated. More concisely, IE p = p, that is,
on the average, the estimator is exactly right.
Definition 9.7. Let s(X\ , X2, ... , Xn) be a statistic which estimates 8. If
IE s(XuX2,...,Xn) = 8,
then the statistic s(X i,X2, . . . ,X„) is said to be an unbiased estimator of 8. Otherwise, it is biased.
212
CHAPTER 9. ESTIMATION
Example 9.8. Let X\, X2, . . . , Xn be an SRS ( n ) from a norm(mean = p, sd = <x) distribution. It
can be shown (in Exercise 9.1) that if 6 = (p, cr2) then the MLE of 6 is
9 = (ft,fr~).
where ft - X and
— \2 n - 1
-.S'2.
We of course know from 8.2 that ft is unbiased. What about cr2? Let us check:
IE cr2 = IE ^ S 2
= IE
cr 2 ( n - 1 )S‘
— IE chisq(df = n - 1)
n
cr2
— (« - 1),
n
from which we may conclude two things:
1 . cr2 is a biased estimator of cr2, and
2. S2 — ncr2/(n - 1) is an unbiased estimator of cr2.
(9.1.3)
(9.1.4)
One of the most common questions in an introductory statistics class is, “Why do we divide
by n - 1 when we compute the sample variance? Why do we not divide by n?” We see now that
division by n amounts to the use of a biased estimator for cr2, that is, if we divided by n then on
the average we would underestimate the true value of cr2. We use n - 1 so that, on the average, our
estimator of cr2 will be exactly right.
9.1.1 How to do it with R
R can be used to find maximum likelihood estimators in a lot of diverse settings. We will discuss
only the most basic here and will leave the rest to more sophisticated texts.
For one parameter estimation problems we may use the optimize function to find MLEs. The
arguments are the function to be maximized (the likelihood function), the range over which the
optimization is to take place, and optionally any other arguments to be passed to the likelihood if
needed.
Let us see how to do Example 9.2. Recall that our likelihood function was given by
L(p)=pLxV -p)n~lx‘. (9.1.5)
Notice that the likelihood is just a product of binom(size = 1, prob = p) PMFs. We first give
some sample data (in the vector datavals), next we define the likelihood function L, and finally
we optimize L over the range c(® , 1).
9.1. POINT ESTIMATION
213
> x <- mtcars$am
> L <- functionCp, x ) prod(dbinom(x, size = 1, prob = p))
> optimize(L , interval = c( 6>, 1) , x - x, maximum = TRUE )
Smaximum
[1] 0.4062458
Sob j ective
[1] 4 . 099989e-10
Note that the optimize function by default minimizes the function L, so we have to set
maximum = TRUE to get an MLE. The returned value of Smaximum gives an approximate value
of the MLE to be 0.406 and $obj ective gives L evaluated at the MLE which is approximately 0.
We previously remarked that it is usually more numerically convenient to maximize the log-
likelihood (or minimize the negative log-likelihood), and we can just as easily do this with R. We
just need to calculate the log-likelihood beforehand which (for this example) is
It is done in R with
> minuslogL <- functionCp, x) -sum(dbinom(x, size = 1, prob = p,
+ log = TRUE))
> optimize (minuslogL, interval = c(&, 1) , x - x)
Sminimum
[1] 0.4062525
Sobjective
[1] 21.61487
Note that we did not need maximum = TRUE because we minimized the negative log-likelihood.
The answer for the MLE is essentially the same as before, but the $ob j ective value was different,
of course.
For multiparameter problems we may use a similar approach by way of the mle function in the
stats4 package.
Example 9.9. Plant Growth. We will investigate the weight variable of the PlantGrowth
data. We will suppose that the weights constitute a random observations X\, W,- • ■ , Xn that are
i.i.d. norm(mean - p, sd - cr ) which is not unreasonable based on a histogram and other ex-
ploratory measures. We will find the MLE of 6 = (ju,<x2). We claimed in Example 9.8 that
6 = (ju, if2) had the form given above. Let us check whether this is plausible numerically. The
negative log-likelihood function is
> minuslogL <- function (mu, sigma2){
+ -sum(dnorm(x, mean = mu, sd = sqrt(sigma2) , log = TRUE))
+ }
214
CHAPTER 9. ESTIMATION
Note that we omitted the data as an argument to the log-likelihood function; the only arguments
were the parameters over which the maximization is to take place. Now we will simulate some
data and find the MLE. The optimization algorithm requires starting values (intelligent guesses)
for the parameters. We choose values close to the sample mean and variance (which turn out to be
approximately 5 and 0.5, respectively) to illustrate the procedure.
> x <- PlantGrowthSweight
> library(stats4)
> MaxLikeEst <- mle(minuslogL , start = list(mu = 5, sigma2 = 6.5))
> summary (MaxLikeEst)
Maximum likelihood estimation
Call:
mle(minuslogl = minuslogL, start = list(mu = 5, sigma2 = 0.5))
Coefficients :
Estimate Std. Error
mu 5.0729848 0.1258666
sigma2 0.4752721 0.1227108
-2 log L: 62.82084
The outputted MLEs are shown above, and mle even gives us estimates for the standard errors
of ji and cf2 (which were obtained by inverting the numerical Hessian matrix at the optima; see
Appendix E.6). Let us check how close the numerical MLEs came to the theoretical MLEs:
> mean(x)
[1] 5.073
> var(x) * 29/36
[1] 0.475281
> sd(x)/sqrt(36)
[1] 0.1280195
The numerical MLEs were very close to the theoretical MLEs. We already knew that the
standard error of ji = X is cr/ yfn, and the numerical estimate of this was very close too.
There is functionality in the distrTest package [74] to calculate theoretical MLEs; we will
skip examples of these for the time being.
9.2 Confidence Intervals for Means
We are given X\, W, •••, X„ that are an SRS(n) from a norm(mean = g, sd = cr) distribution,
where g is unknown. We know that we may estimate g with X, and we have seen that this estimator
is the MLE. But how good is our estimate? We know that
X-Ll
— ~ norm(mean = 0, sd = 1).
cr! -y n
(9.2.1)
9.2. CONFIDENCE INTERVALS FOR MEANS
215
For a big probability 1 - a, for instance, 95%, we can calculate the quantile za/2- Then
P
X-
~Za/2
cr / \[n
: Za/ 2
= 1
But now consider the following string of equivalent inequalities:
(9.2.2)
~Za/2 ^
x-fi
cr/ yfn
^ Za/2,
Za/2l^p| <X-H <Zff/2(-^|,
\fn
\ V«
- Za/2 ( — p I < -^ < -X + Za/2 I ~j= I »
\
a/ Ti
X - Za/2
x + z<*n ~~f •
cr
That is,
P(x-za/ 2-^= <yU < X + Za/2-^J =l-a.
Definition 9.10. The interval
— cr — cr
X - Za/2 —p , X + Za/2 ~p
yn yn
(9.2.3)
(9.2.4)
is a 100(1 - a)% confidence interval for fi. The quantity 1 - a is called the confidence coefficient.
Remark 9. 1 1. The interval is also sometimes written more compactly as
(9.2.5)
— <T
X ± Zal2 ■
-fn
The interpretation of confidence intervals is tricky and often mistaken by novices. When I am
teaching the concept “live” during class, I usually ask the students to imagine that my piece of
chalk represents the “unknown” parameter, and I lay it down on the desk in front of me. Once the
chalk has been lain, it is fixed; it does not move. Our goal is to estimate the parameter. For the
estimator I pick up a sheet of loose paper lying nearby. The estimation procedure is to randomly
drop the piece of paper from above, and observe where it lands. If the piece of paper covers the
piece of chalk, then we are successful - our estimator covers the parameter. If it falls off to one side
or the other, then we are unsuccessful; our interval fails to cover the parameter.
Then I ask them: suppose we were to repeat this procedure hundreds, thousands, millions of
times. Suppose we kept track of how many times we covered and how many times we did not.
What percentage of the time would we be successful?
In the demonstration, the parameter corresponds to the chalk, the sheet of paper corresponds
to the confidence interval, and the random experiment corresponds to dropping the sheet of paper.
The percentage of the time that we are successful exactly corresponds to the confidence coefficient.
That is, if we use a 95% confidence interval, then we can say that, in the long run, approximately
95% of our intervals will cover the true parameter (which is fixed, but unknown).
216
CHAPTER 9. ESTIMATION
Confidence intervals based on z distribution
X
o
T3
c:
o
LO
O
o
CO
o
C\J
o -
95
100
105
Confidence Interval
Figure 9.2.1: Simulated confidence intervals
The graph was generated by the ci . examp function from the TeachingDemos package. Fifty (50) samples
of size twenty five (25) were generated from a norm(mean = 100, sd = 10) distribution, and each sample
was used to find a 95% confidence interval for the population mean using Equation 9.2.5. The 50 confidence
intervals are represented above by horizontal lines, and the respective sample means are denoted by vertical
slashes. Confidence intervals that “cover” the true mean value of 100 are plotted in black; those that fail to
cover are plotted in a lighter color. In the plot we see that only one (1) of the simulated intervals out of the 50
failed to cover p = 100, which is a success rate of 98%. If the number of generated samples were to increase
from 50 to 500 to 50000, . . . , then we would expect our success rate to approach the exact value of 95%.
9.2. CONFIDENCE INTERVALS FOR MEANS
217
See Figure 9.2.1, which is a graphical display of these ideas.
Under the above framework, we can reason that an “interval” with a larger confidence coef-
ficient corresponds to a wider sheet of paper. Furthermore, the width of the confidence interval
(sheet of paper) should be somehow related to the amount of information contained in the random
sample, X\, W>, . . . , X„. The following remarks makes these notions precise.
Remark 9.12. For a fixed confidence coefficient 1 - a,
if n increases, then the confidence interval gets SHORTER. (9.2.6)
Remark 9. 13. For a fixed sample size n,
if 1 - a increases, then the confidence interval gets WIDER. (9.2.7)
Example 9.14. Results from an Experiment on Plant Growth. The PlantGrowth data frame
gives the results of an experiment to measure plant yield (as measured by the weight of the plant).
We would like to a 95% confidence interval for the mean weight of the plants. Suppose that we
know from prior research that the true population standard deviation of the plant weights is 0.7 g.
The parameter of interest is /j, which represents the true mean weight of the population of all
plants of the particular species in the study. We will first take a look at a stemplot of the data:
> library (aplpack)
> with(PlantGrowth, stem. leaf (weight))
1 | 2: represents 1.2
leaf unit : 0.1
n: 30
1
f
1
5
s
1
2
3.
1
8
4
4*
1
11
5
t
1
3
8
f
1
455
1®
s
1
66
13
4.
1
889
(4)
5*
1
1111
13
t
1
2233
9
f
1
555
s
1
6
5.
1
88
4
6*
1
011
1
t
1
3
The data appear to be approximately normal with no extreme values. The data come from a
designed experiment, so it is reasonable to suppose that the observations constitute a simple random
sample of weights3. We know the population standard deviation cr = 0.70 from prior research. We
are going to use the one-sample z- interval.
3 Actually we will see later that there is reason to believe that the observations are simple random samples from three
distinct populations. See Section 10.6.
218
CHAPTER 9. ESTIMATION
> dim(PlantGrowth) # sample size is first entry
[1] 30 2
> with(PlantGrowth, mean(weight))
[1] 5.073
> qnorm(& . 975)
[1] 1.959964
We find the sample mean of the data to be x = 5.073 and za/2 = Z0.025 ~ 1-96. Our interval is
therefore
cr 0.70
x ± Za/2 — -p = 5.073 ± 1.96 ■ — — ,
y/n V30
which comes out to approximately [4.823, 5.323], In conclusion, we are 95% confident that the
true mean weight p of all plants of this species lies somewhere between 4.823 g and 5.323 g, that
is, we are 95% confident that the interval [4.823, 5.323] covers p. See Figure
Example 9.15. Give some data with X\, X2, . . . , X„ an SRS (n) from a normfmean = p, sd = cr)
distribution. Maybe small sample?
1. What is the parameter of interest? in the context of the problem. Give a point estimate for p.
2. What are the assumptions being made in the problem? Do they meet the conditions of the
interval?
3. Calculate the interval.
4. Draw the conclusion.
Remark 9. 16. What if cr is unknown? We instead use the interval
X±Za/2~p- (9.2.8)
yn
where S is the sample standard deviation.
• If n is large, then X will have an approximately normal distribution regardless of the under-
lying population (by the CLT) and S will be very close to the parameter cr (by the SLLN);
thus the above interval will have approximately 100(1 - a)% confidence of covering p.
• If n is small, then
o If the underlying population is normal then we may replace za/2 with ta/2 (df = n — 1).
The resulting 100(1 - a)% confidence interval is
X±ta/2(df = n-l)4= (9-2.9)
o if the underlying population is not normal, but approximately normal, then we may
use the t interval. Equation 9.2.9. The interval will have approximately 100(1 - a)%
confidence of covering p. However, if the population is highly skewed or the data have
outliers, then we should ask a professional statistician for advice.
9.2. CONFIDENCE INTERVALS FOR MEANS
219
(I X
95% Normal Confidence Limits: o* = 0.128, n = 30
5.073
-1.96
1-96 Conf Level= 0.9500
Figure 9.2.2: Confidence interval plot for the PlantGrowth data
The shaded portion represents 95% of the total area under the curve, and the upper and lower bounds are the
limits of the one-sample 95% confidence interval. The graph is centered at the observed sample mean. It was
generated by computing a z . test from the TeachingDemos package, storing the resulting htest object, and
plotting it with the normal . and . t . dist function from the HH package. See the remarks in the “How to do it
with FT" discussion later in this section.
220
CHAPTER 9. ESTIMATION
The author learned of a handy acronym from AP Statistics Exam graders that summarizes the
important parts of confidence interval estimation, which is PANIC: Parameter, Assumptions, Name,
Interval, and Conclusion.
Parameter: identify the parameter of interest with the proper symbols. Write down what the
parameter means in the context of the problem.
Assumptions: list any assumptions made in the experiment. If there are any other assumptions
needed or that were not checked, state what they are and why they are important.
Name: choose a statistical procedure from your bag of tricks based on the answers to the previous
two parts. The assumptions of the procedure you choose should match those of the problem;
if they do not match then either pick a different procedure or openly admit that the results
may not be reliable. Write down any underlying formulas used.
Interval: calculate the interval from the sample data. This can be done by hand but will more
often be done with the aid of a computer. Regardless of the method, all calculations or code
should be shown so that the entire process is repeatable by a subsequent reader.
Conclusion: state the final results, using language in the context of the problem. Include the
appropriate interpretation of the interval, making reference to the confidence coefficient.
Remark 9. 17. All of the above intervals for ft were two-sided, but there are also one-sided intervals
for ft. They look like
cr
X-za—=, °°
yn
cr
X + Za—p
yn
(9.2.10)
and satisfy
P - za~^p < fij - 1 - a and P > fi | = 1 - a.
(9.2.11)
Example 9.18. Small sample, some data with Xt, X2, . .., Xn an SRS(n) from a norm(mean =
H, sd = cr) distribution.
1. PANIC
9.2.1 How to do it with R
We can do Example 9.14 with the following code.
> library (TeachingDemos)
> temp <- with(PlantGrowth, z . test (weight , stdev = 6.7))
> temp
One Sample z-test
data: weight
z = 39.6942, n = 30.88®, Std. Dev. = 8.700, Std. Dev. of the
sample mean = 8.128, p-value < 2.2e-16
alternative hypothesis: true mean is not equal to 0
95 percent confidence interval:
9.3. CONFIDENCE INTERVALS FOR DIFFERENCES OF MEANS
221
4.822513 5.323487
sample estimates:
mean of weight
5.873
The confidence interval bounds are shown in the sixth line down of the output (please disregard
all of the additional output information for now - we will use it in Chapter 10). We can make the
plot for Figure 9.2.2 with
> library (IPSUR)
> plot (temp, "Conf")
9.3 Confidence Intervals for Differences of Means
Let X\, X2, . . . , Xn be a SRS (n) from a norm(mean = px, sd = crx) distribution and let Y\. Yn, . . . ,
Ym be a SRS (in) from a norm(mean = py, sd = cry) distribution. Further, assume that the X\. Xo,
X„ sample is independent of the F1; Y2, . . . , Ym sample.
Suppose that crx and cry are known. We would like a confidence interval for px - py. We know
Unfortunately, most of the time the values of crx and cry are unknown. This leads us to the follow-
ing:
• If both sample sizes are large, then we may appeal to the CLT/SLLN (see 8.3) and substitute
S 2 and Sy for cr2x and cr2 in the interval 9.3.2. The resulting confidence interval will have
approximately 100(1 - a)% confidence.
• If one or more of the sample sizes is small then we are in trouble, unless
o the underlying populations are both normal and crx = cry. In this case (setting cr =
crx = cry).
that
(9.3.1)
Therefore, a 100(1 - a)% confidence interval for px - py is given by
(9.3.2)
(9.3.3)
Now let
(9.3.4)
Then by Exercise 7.2 we know that U ~ chisqldf = n + m - 2) and is not a large leap
to believe that U is independent of X - Y; thus
T = ~ t(df = n + m - 2).
T =
V U / (n + m — 2)
~ t(df = n + m - 2).
(9.3.5)
222
CHAPTER 9. ESTIMATION
But
X-Y-(J1X-Ily)
T =
X -Y - (px ~ Py)
V
~ t(df = n + m - 2).
Therefore a 100(1 - a)% confidence interval for px - py is given by
(x - y'j + taji( df = n + m - 2 )S p
1 1
n m ’
(9.3.6)
where
(n - 1)S|. + (m - 1)5 y
n +IH — 2
(9.3.7)
is called the “pooled” estimator of cr.
o if one of the samples is small, and both underlying populations are normal, but crx 4-
crY , then we may use a Welch (or Satterthwaite) approximation to the degrees of free-
dom. See Welch [88], Satterthwaite [76], or Neter el al [67]. The idea is to use an
interval of the form
where the degrees of freedom r is chosen so that the interval has nice statistical proper-
ties. It turns out that a good choice for r is given by
where we understand that r is rounded down to the nearest integer. The resulting inter-
val has approximately 100(1 - a)% confidence.
9.3.1 How to do it with R
The basic function is t . test which has a var . equal argument that may be set to TRUE or FALSE.
The confidence interval is shown as part of the output, although there is a lot of additional informa-
tion that is not needed until Chapter 10.
There is not any specific functionality to handle the z-interval for small samples, but if the
samples are large then t . test with var . equal = FALSE will be essentially the same thing. The
standard deviations are never (?) known in advance anyway so it does not really matter in practice.
(9.3.8)
r =
(9.3.9)
9.4. CONFIDENCE INTERVALS FOR PROPORTIONS
223
9.4 Confidence Intervals for Proportions
We would like to know p which is the “proportion of successes”. For instance, p could be:
• the proportion of U.S. citizens that support Obama,
• the proportion of smokers among adults age 1 8 or over,
• the proportion of people worldwide infected by the H1N1 virus.
We are given an SRS(n) X\, Xi, ..., X„ distributed binom(size = 1, prob = p). Recall from
Section 5.3 that the common mean of these variables is E X = p and the variance is 1E(X - p)2 =
p( 1 - p). If we let Y = 2 Xi , then from Section 5.3 we know that Y ~ binom(size = n, prob = p)
and that
— Y — - p(l-p)
X = — has IEX = p and Var(V) = — — .
n n
Thus if n is large (here is the CLT) then an approximate 100(1 - a)% confidence interval for p
would be given by
V + z,
■a/2
p(l-p)
(9.4.1)
OOPS. . . ! Equation 9.4.1 is of no use to us because the unknown parameter p is in the formula!
(If we knew what p was to plug in the formula then we would not need a confidence interval in the
first place.) There are two solutions to this problem.
1. Replace p with p — X. Then an approximate 100(1 - a)% confidence interval for p is given
by
P ± Za/ 2
PO--P)
(9.4.2)
This approach is called the Wald interval and is also known as the asymptotic interval because
it appeals to the CLT for large sample sizes.
2. Go back to first principles. Note that
-Za/ 2 <
Y/n - p
VpO - P)/n
exactly when the function f defined by
f(p) = C Y/n - p)
: Za/2
2 2 P(\~P)
Z-a/2 n
satisfies f(p) < 0. But / is quadratic in p so its graph is a parabola; it has two roots, and
these roots form the limits of the confidence interval. We can find them with the quadratic
formula (see Exercise 9.2):
P +
'■a/2
2 n
± Za/2 '
Pd-P) , Z*/2
(2 n)2
'■a/2
n
(9.4.3)
This approach is called the score inten’al because it is based on the inversion of the “Score
test”. See Chapter 14. It is also known as the Wilson interval, see Agresti [3].
224
CHAPTER 9. ESTIMATION
For two proportions p\ and p2, we may collect independent binom(size = 1, prob = p) samples
of size n \ and ri2, respectively. Let Y\ and Yi denote the number of successes in the respective
samples.
We know that
and
Y\
— norm
n\
mean = px, sd =
— norm
«2
mean = p2, sd =
so it stands to reason that an approximate 100(1 - a)% confidence interval for p\ - p2 is given by
(P t - Pi) ± za/ 2
P id - Pi)
n i
/32(1 ~ pi)
«2
(9.4.4)
where p\ = Y\/n\ and /L = L2/«2-
Remark 9.19. When estimating a single proportion, one-sided intervals are sometimes needed.
They take the form
0, p + Za/2
P(l-P)
(9.4.5)
or
P ~ Za/2
PO -P)
, 1
(9.4.6)
or in other words, we know in advance that the true proportion is restricted to the interval [0,1], so
we can truncate our confidence interval to those values on either side.
9.4.1 How to do it with R
> library (Hmisc)
> binconf(x = 7, n = 25, method = "asymptotic")
PointEst Lower Upper
0.28 0.1039957 0.4560043
> binconfCx = 7, n = 25, method = "wilson")
PointEst Lower Upper
0.28 0.1428385 0.4757661
The default value of the method argument is wilson.
An alternate way is
> tab <- xtabs(~gender, data = RcmdrTestDrive)
> prop. test (rbind(tab) , conf. level = 6.95, correct = FALSE)
9.5. CONFIDENCE INTERVALS FOR VARIANCES
225
1-sample proportions test without continuity correction
data: rbind(tab) , null probability 0.5
X-squared = 2.881, df = 1 , p-value = 0.08963
alternative hypothesis: true p is not equal to 0.5
95 percent confidence interval:
0.4898844 0.6381406
sample estimates:
P
0.5654762
> A <- as . data . frame (Titanic)
> library (reshape)
> B <- with(A, untable(A, Freq))
9.5 Confidence Intervals for Variances
I am thinking one and two sample problems here.
9.5.1 How to do it with R
I am thinking about sigma. test in the TeachingDemos package and var.test in base R here.
9.6 Fitting Distributions
9.6.1 How to do it with R
I am thinking about fitdistr from the MASS package [84].
9.7 Sample Size and Margin of Error
Sections 9.2 through 9.5 all began the same way: we were given the sample size n and the confi-
dence coefficient 1 - a, and our task was to find a margin of error E so that
8 ± E is a 100(1 - a)% confidence interval for 8.
Some examples we saw were:
• E = Zaiicrl yfn, in the one-sample z- interval,
• E - fff/2(df = n + m - 2 )S p Vfr1 + m~l, in the two-sample pooled r-interval.
226
CHAPTER 9. ESTIMATION
We already know (we can see in the formulas above) that E decreases as n increases. Now we
would like to use this information to our advantage: suppose that we have a fixed margin of error
E, say E - 3, and we want a 100(1 - a)% confidence interval for p. The question is: how big does
n have to be?
For the case of a population mean the answer is easy: we set up an equation and solve for n.
Example 9.20. Given a situation, given <r, given E, we would like to know how big n has to be to
ensure that X ± 5 is a 95% confidence interval for p.
Remark 9.21.
1. Always round up any decimal values of n, no matter how small the decimal is.
2. Another name for E is the “maximum error of the estimate”.
For proportions, recall that the asymptotic formula to estimate p was
P ± Za/2
pa-p)
Reasoning as above we would want
E = z,
■a/2
pa-P)
or
_2 P( 1 “ P)
W 2 £ 2
(9.7.1)
(9.7.2)
OOPS ! Recall that p = Y/n, which would put the variable n on both sides of Equation 9.7.2. Again,
there are two solutions to the problem.
1. If we have a good idea of what p is, say p* then we can plug it in to get
2 P'd-P*)
‘-a/ 2
E2
(9.7.3)
2.
Even if we have no idea what p is, we do know from calculus that p( \ —p)< 1/4 because the
function f(x) - x( 1 - x) is quadratic (so its graph is a parabola which opens downward) with
maximum value attained at x = 1 /2. Therefore, regardless of our choice for p* the sample
size must satisfy
2 P*d-P‘) . Zl/2
l~Za'2 E 2 ~ 4E2 '
(9.7.4)
The quantity z2^/^E2 is large enough to guarantee 100(1 - a)% confidence.
Example 9.22. Proportion example
Remark 9.23. For very small populations sometimes the value of n obtained from the formula is
too big. In this case we should use the hypergeometric distribution for a sampling model rather than
the binomial model. With this modification the formulas change to the following: if N denotes the
population size then let
m = z
2
a/2
P*( 1 - Pi
E2
(9.7.5)
9.8. OTHER TOPICS
227
and the sample size needed to ensure 100(1 - a)% confidence is achieved is
m
If we do not have a good value for the estimate p* then we may use p* — 1 /2.
(9.7.6)
9.7.1 How to do it with R
I am thinking about power.t.test, power.prop.test, power.anova.test, and I am also thinking about
replicate.
9.8 Other Topics
Mention mle from the stats4 package.
228
CHAPTER 9. ESTIMATION
Chapter Exercises
Exercise 9.1. Let X\, AS, . . . , Xn be an S RS (n) from a norm(raean = sd = cr) distribution. Find
a two-dimensional MLE for 6 — (p, cr).
Exercise 9.2. Find the upper and lower limits for the confidence interval procedure by finding the
roots of / defined by
,, s ,v/ \2 2 P(l ~P)
f(P ) = (Y/n - p) - za,2 •
' n
You are going to need the quadratic formula.
Chapter 10
Hypothesis Testing
What do I want them to know?
• basic terminology and philosophy of the Neyman-Pearson paradigm
• classical hypothesis tests for the standard one and two sample problems with means, vari-
ances, and proportions
• the notion of between versus within group variation and how it plays out with one-way
ANOVA
• the concept of statistical power and its relation to sample size
10.1 Introduction
I spent a week during the summer of 2005 at the University of Nebraska at Lincoln grading Ad-
vanced Placement Statistics exams, and while I was there I attended a presentation by Dr. Roxy
Peck. At the end of her talk she described an activity she had used with students to introduce the
basic concepts of hypothesis testing. I was impressed by the activity and have used it in my own
classes several times since.
The instructor (with a box of cookies in hand) enters a class of fifteen or more students and
produces a brand-new, sealed deck of ordinary playing cards. The instructor asks for a student
volunteer to break the seal, and then the instructor prominently shuffles the deck1 several times in
front of the class, after which time the students are asked to line up in a row. They are going to
play a game. Each student will draw a card from the top of the deck, in turn. If the card is black,
then the lucky student will get a cookie. If the card is red, then the unlucky student will sit down
empty-handed. Let the game begin.
The first student draws a card: red. There are jeers and outbursts, and the student slinks off to
his/her chair. (S)he is disappointed, of course, but not really. After all, (s)he had a 50-50 chance of
getting black, and it did not happen. Oh well.
'The jokers are removed before shuffling.
229
230
CHAPTER 10. HYPOTHESIS TESTING
The second student draws a card: red, again. There are more jeers, and the second student
slips away. This student is also disappointed, but again, not so much, because it is probably his/her
unlucky day. On to the next student.
The student draws: red again! There are a few wiseguys who yell (happy to make noise, more
than anything else), but there are a few other students who are not yelling any more - they are
thinking. This is the third red in a row, which is possible, of course, but what is going on, here?
They are not quite sure. They are now concentrating on the next card. . . it is bound to be black,
right?
The fourth student draws: red. Hmmm. . . now there are groans instead of outbursts. A few of
the students at the end of the line shrug their shoulders and start to make their way back to their
desk, complaining that the teacher does not want to give away any cookies. There are still some
students in line though, salivating, waiting for the inevitable black to appear.
The fifth student draws red. Now it isn’t funny any more. As the remaining students make their
way back to their seats an uproar ensues, from an entire classroom demanding cookies.
Keep the preceding experiment in the back of your mind as you read the following sections.
When you have finished the entire chapter, come back and read this introduction again. All of the
mathematical jargon that follows is connected to the above paragraphs. In the meantime, I will get
you started:
Null hypothesis: it is an ordinary deck of playing cards, shuffled thoroughly.
Alternative hypothesis: either it is a trick deck of cards, or the instructor did some fancy shuffle-
work.
Observed data: a sequence of draws from the deck, five reds in a row.
If it were truly an ordinary, well-shuffled deck of cards, the probability of observing zero blacks
out of a sample of size five (without replacement) from a deck with 26 black cards and 26 red cards
would be
> dhyper(Q, m = 26, n = 26, k = 5)
[1] O.Q2531Q12
There are two very important final thoughts. First, everybody gets a cookie in the end. Second,
the students invariably (and aggressively) attempt to get me to open up the deck and reveal the true
nature of the cards. I never do.
10.2 Tests for Proportions
Example 10.1. We have a machine that makes widgets.
• Under normal operation, about 0. 10 of the widgets produced are defective.
• Go out and purchase a torque converter.
• Install the torque converter, and observe n = 100 widgets from the machine.
• Let Y - number of defective widgets observed.
10.2. TESTS FOR PROPORTIONS
231
If
• Y = 0, then the torque converter is great!
• Y = 4, then the torque converter seems to be helping.
• Y = 9, then there is not much evidence that the torque converter helps.
• Y — 17, then throw away the torque converter.
Let p denote the proportion of defectives produced by the machine. Before the installation of the
torque converter p was 0.10. Then we installed the torque converter. Did p change? Did it go up or
down? We use statistics to decide. Our method is to observe data and construct a 95% confidence
interval for p,
If the confidence interval is
• [0.01, 0.05], then we are 95% confident that 0.01 < p < 0.05, so there is evidence that the
torque converter is helping.
• [0.15, 0.19], then we are 95% confident that 0.15 < p < 0.19, so there is evidence that the
torque converter is hurting.
• [0.07, 0.1 1], then there is not enough evidence to conclude that the torque converter is doing
anything at all, positive or negative.
10.2.1 Terminology
The null hypothesis Hq is a “nothing” hypothesis, whose interpretation could be that nothing has
changed, there is no difference, there is nothing special taking place, etc.. In Example 10.1 the
null hypothesis would be Ho : p — 0.10. The alternative hypothesis H i is the hypothesis that
something has changed, in this case, H\ : p + 0.10. Our goal is to statistically test the hypothesis
//o : p — 0.10 versus the alternative H\ \ p + 0.10. Our procedure will be:
1 . Go out and collect some data, in particular, a simple random sample of observations from the
machine.
2. Suppose that Ho is true and construct a 100(1 - a)% confidence interval for p.
3. If the confidence interval does not cover p = 0.10, then we rejectHo- Otherwise, we fail to
rejectHo-
Remark 10.2. Every time we make a decision it is possible to be wrong, and there are two possible
mistakes that we could make. We have committed a
Type I Error if we reject Ho when in fact Hq is true. This would be akin to convicting an innocent
person for a crime (s)he did not commit.
(10.2.1)
Type II Error if we fail to reject Hq when in fact If is true. This is analogous to a guilty person
escaping conviction.
232
CHAPTER 10. HYPOTHESIS TESTING
Type I Errors are usually considered worse2, and we design our statistical procedures to control the
probability of making such a mistake. We define the
significance level of the test = P(Type I Error) = a. (10.2.2)
We want a to be small which conventionally means, say, a = 0.05, a = 0.01, or a = 0.005 (but
could mean anything, in principle).
• The rejection region (also known as the critical region) for the test is the set of sample values
which would result in the rejection of Hq. For Example 10.1, the rejection region would be
all possible samples that result in a 95% confidence interval that does not cover p = 0.10.
• The above example with H\ : p + 0.10 is called a two-sided test. Many times we are
interested in a one-sided test, which would look like H\ : p < 0.10 or H\ : p > 0.10.
We are ready for tests of hypotheses for one proportion.
Table here.
Don’t forget the assumptions.
Example 10.3. Find
1 . The null and alternative hypotheses
2. Check your assumptions.
3. Define a critical region with an a — 0.05 significance level.
4. Calculate the value of the test statistic and state your conclusion.
Example 10.4. Suppose p = the proportion of students who are admitted to the graduate school
of the University of California at Berkeley, and suppose that a public relations officer boasts that
UCB has historically had a 40% acceptance rate for its graduate school. Consider the data stored in
the table UCBAdmissions from 1973. Assuming these observations constituted a simple random
sample, are they consistent with the officer’s claim, or do they provide evidence that the acceptance
rate was significantly less than 40%? Use an a - 0.01 significance level.
Our null hypothesis in this problem is H$ : p = 0.4 and the alternative hypothesis is H\ : p <
0.4. We reject the null hypothesis if p is too small, that is, if
p - 0.4
V0.4(l - 0.4)/n
< - Za ,
(10.2.3)
where a — 0.01 and -Zo.oi is
> -qnorm((S . 99)
[1] -2.326348
-There is no mathematical difference between the errors, however. The bottom line is that we choose one type of
error to control with an iron fist, and we try to minimize the probability of making the other type. That being said, null
hypotheses are often by design to correspond to the “simpler” model, so it is often easier to analyze (and thereby control)
the probabilities associated with Type I Errors.
10.2. TESTS FOR PROPORTIONS
233
Our only remaining task is to find the value of the test statistic and see where it falls relative to
the critical value. We can find the number of people admitted and not admitted to the UCB graduate
school with the following.
> A <- as. data, frame (UCBAdmissions)
> head (A)
Admit
Gender Dept
Freq
1
Admitted
Male
A
512
2
Rejected
Male
A
313
3
Admitted
Female
A
89
4
Rejected
Female
A
19
5
Admitted
Male
B
353
6
Rejected
Male
B
207
>
xtabs(Freq ~ Admit
, data ■
Admit
Admitted Rejected
1755 2771
Now we calculate the value of the test statistic.
> phat <- 1755/(1755 + 2771)
> (phat - ®. 4)/sqrt (0. 4 * 6.6/(1755 + 2771))
[1] -1.680919
Our test statistic is not less than -2.32, so it does not fall into the critical region. Therefore, we
fail to reject the null hypothesis that the true proportion of students admitted to graduate school is
less than 40% and say that the observed data are consistent with the officer’s claim at the a = 0.01
significance level.
Example 10.5. We are going to do Example 10.4 all over again. Everything will be exactly the
same except for one change. Suppose we choose significance level a = 0.05 instead of a — 0.01.
Are the 1973 data consistent with the officer’s claim?
Our null and alternative hypotheses are the same. Our observed test statistic is the same: it was
approximately -1.68. But notice that our critical value has changed: a = 0.05 and -zo.os is
> -qnorm(6 . 95)
[1] -1.644854
Our test statistic is less than -1.64 so it now falls into the critical region! We now reject the null
hypothesis and conclude that the 1973 data provide evidence that the true proportion of students
admitted to the graduate school of UCB in 1973 was significantly less than 40%. The data are not
consistent with the officer’s claim at the a = 0.05 significance level.
What is going on, here? If we choose a = 0.05 then we reject the null hypothesis, but if we
choose a = 0.01 then we fail to reject the null hypothesis. Our final conclusion seems to depend
on our selection of the significance level. This is bad; for a particular test, we never know whether
our conclusion would have been different if we had chosen a different significance level.
234
CHAPTER 10. HYPOTHESIS TESTING
Or do we?
Clearly, for some significance levels we reject, and for some significance levels we do not.
Where is the boundary? That is, what is the significance level for which we would reject at any sig-
nificance level bigger, and we would fail to reject at any significance level smaller ? This boundary
value has a special name: it is called the p-value of the test.
Definition 10.6. The p-value, or observed significance level, of a hypothesis test is the probability
when the null hypothesis is true of obtaining the observed value of the test statistic (such as p) or
values more extreme - meaning, in the direction of the alternative hypothesis3.
Example 10.7. Calculate the p-value for the test in Examples 10.4 and 10.5.
The p-value for this test is the probability of obtaining a z-score equal to our observed test
statistic (which had ’-score ~ -1.680919) or more extreme, which in this example is less than the
observed test statistic. In other words, we want to know the area under a standard normal curve on
the interval (-oo, -1.680919]. We can get this easily with
> pnorm(- 1.680919)
[1] 0.84638932
We see that the p-value is strictly between the significance levels a = 0.01 and a = 0.05.
This makes sense: it has to be bigger than a = 0.01 (otherwise we would have rejected Ho in
Example 10.4) and it must also be smaller than a = 0.05 (otherwise we would not have rejected
Ho in Example 10.5). Indeed, p-values are a characteristic indicator of whether or not we would
have rejected at assorted significance levels, and for this reason a statistician will often skip the
calculation of critical regions and critical values entirely. If (s)he knows the p-value, then (s)he
knows immediately whether or not (s)he would have rejected at any given significance level.
Thus, another way to phrase our significance test procedure is: we will reject Ho at the ar-level
of significance if the p-value is less than a.
Remark 10.8. If we have two populations with proportions p\ and p2 then we can test the null
hypothesis H0 : pi = p2-
Table Here.
Example 10.9. Example.
10.2.2 How to do it with R
The following does the test.
> prop . test(1755 , 1755 + 2771, p = 0.4, alternative = "less",
+ conf. level = 0.99, correct = FALSE)
1-sample proportions test without continuity correction
data: 1755 out of 1755 + 2771, null probability 0.4
X-squared = 2.8255, df = 1, p-value = 0.04639
'Bickel and Doksum [7] state the definition particularly well: the p-value is “the smallest level of significance a at which
an experimenter using [the test statistic] T would reject [//o] on the basis of the observed [sample] outcome x”.
10.3. ONE SAMPLE TESTS FOR MEANS AND VARIANCES
235
alternative hypothesis: true p is less than 0.4
99 percent confidence interval:
0.0000000 0.4047326
sample estimates:
P
0.3877596
Do the following to make the plot.
> library (IPSUR)
> library (HH)
> temp <- prop ,test(1755 , 1755 + 2771, p = 0.4, alternative = "less",
+ conf. level = 0.99, correct = FALSE)
> plot (temp, "Hypoth")
Use Yates’ continuity correction when the expected frequency of successes is less than 10. You
can use it all of the time, but you will have a decrease in power. For large samples the correction
does not matter.
With the R Commander If you already know the number of successes and failures, then you
can use the menu Statistics > Proportions > IPSUR Enter table for single sample. . .
Otherwise, your data - the raw successes and failures - should be in a column of the Active
Data Set. Furthermore, the data must be stored as a “factor” internally. If the data are not a factor
but are numeric then you can use the menu Data > Manage variables in active data set > Convert
numeric variables to factors. . . to convert the variable to a factor. Or, you can always use the
factor function.
Once your unsummarized data is a column, then you can use the menu Statistics > Proportions
> Single-sample proportion test. . .
10.3 One Sample Tests for Means and Variances
10.3.1 For Means
Here, X\, Yj, . . . , Xn are a S RS (n) from a normfmean - p, sd = <r) distribution. We would like to
test H0: p= p0.
Case A: Suppose cr is known. Then under Hq,
Z - — — ^ ~ normfmean = 0, sd = 1).
cr/ yn
Table here.
Case B: When cr is unknown, under Hq
X- Po
S / yfn
~ t(df = n- 1).
Table here.
236
CHAPTER 10. HYPOTHESIS TESTING
normal density: 0*= 0.007, n = 1
* 0.383
z -3-2-10123
-2.326
-1.681
a = 0.0100
p = 0.0464
Figure 10.2.1: Hypothesis test plot based on normal . and. t . dist from the HH package
This plot shows all of the important features of hypothesis tests in one magnificent display. The (asymptotic)
distribution of the test statistic (under the null hypothesis) is standard normal, represented by the bell curve,
above. We see the critical region to the left, and the blue shaded area is the significance level, which for this
example is a = 0.05. The area outlined in green is the p-value, and the observed test statistic determines the
upper bound of this region. We can see clearly that the p-value is larger than the significance level, thus, we
would not reject the null hypothesis. There are all sorts of tick marks shown below the graph which detail how
the different pieces are measured on different scales (the original data scale, the standardized scale, etc.). The
workhorse behind the plot is the normal . and . t . dist function from the HH package. See the discussion
in “How to do it with FT' for the exact sequence of commands to generate the plot.
10.3. ONE SAMPLE TESTS FOR MEANS AND VARIANCES
237
Remark 10.10. If u is unknown but n is large then we can use the z-test.
Example 10.11. In this example we
1. Find the null and alternative hypotheses.
2. Choose a test and find the critical region.
3. Calculate the value of the test statistic and state the conclusion.
4. Find the p- value.
Remark 10.12. Remarks
• p- values are also known as tail end probabilities. We reject Hq when the p- value is small.
• cr/ yfn when cr is known, is called the standard error of the sample mean. In general, if we
have an estimator 0 then erg is called the standard error of 0. We usually need to estimate erg
with <j8.
10.3.2 How to do it with R
I am thinking z . test in TeachingDemos, t . test in base R.
> x <- rnorm(37, mean -2, sd = 3)
> library (TeachingDemos)
> z.test(x, mu - 1, sd - 3, conf. level = 8.9)
One Sample z-test
data: x
z = 2.8126, n = 37.888, Std. Dev. = 3.888, Std. Dev. of the sample
mean = 8.493, p-value = 8.884914
alternative hypothesis: true mean is not equal to 1
98 percent confidence interval:
1.575948 3.198422
sample estimates:
mean of x
2.387185
The RcmdrPlugin . IPSUR package does not have a menu for z . test yet.
> x <- rnorm(13, mean -2, sd = 3)
> t.test(x, mu = 8, conf . level = 8.9, alternative = "greater")
One Sample t-test
data: x
t = 1.2949, df = 12, p-value = 8.1899
alternative hypothesis: true mean is greater than 8
98 percent confidence interval:
238
CHAPTER 10. HYPOTHESIS TESTING
normal density: 0*= 0.007, n = 1
* 0.383
z -3 -2
3
-2.326
-1.681
a = 0.0100
p = 0.0464
Figure 10.3.1: Hypothesis test plot based on normal . and. t . dist from the HH package
This plot shows the important features of hypothesis tests.
10.4. TWO-SAMPLE TESTS FOR MEANS AND VARIANCES
239
-0 .05064(306 Inf
sample estimates:
mean of x
1.068850
With the R Commander Your data should be in a single numeric column (a variable) of the
Active Data Set. Use the menu Statistics > Means > Single-sample t-test. . .
10.3.3 Tests for a Variance
Here, X\ , X2, ..., X„ are a S RS (n) from a norm(mean = fi, sd
test //o : cr2 = (To. We know that under Hq,
X2 = — — ^ chisq(df = n
cr-
Table here.
Example 10.13. Give some data and a hypothesis.
1 . Give an cr-level and test the critical region way.
2. Find the p- value for the test.
10.3.4 How to do it with R
I am thinking about sigma . test in the TeachingDemos package.
> library (TeachingDemos)
> sigma. test (womenSheight, sigma = 8 )
One sample Chi-squared test for variance
data: womenSheight
X-squared = 4.375, df = 14, p-value = 0.01449
alternative hypothesis: true variance is not equal to 64
95 percent confidence interval:
10.72019 49.74483
sample estimates:
var of womenSheight
20
10.4 Two-Sample Tests for Means and Variances
The basic idea for this section is the following. We have X ~ norm(mean = jux, sd = <rx) and
Y ~ norm(mean = //>-, sd = cry), distributed independently. We would like to know whether X
and Y come from the same population distribution, that is, we would like to know:
Does A = T?
= cr) distribution. We would like to
-1).
(10.4.1)
240
CHAPTER 10. HYPOTHESIS TESTING
where the symbol = means equality of probability distributions.
Since both X and Y are normal, we may rephrase the question:
Does fix = Py and o"x = cry?
(10.4.2)
Suppose first that we do not know the values of ux and crY, but we know that they are equal,
(Tx = cry. Our test would then simplify to Hq : jjx = fiy. We collect data X \ , W, . . . , Xn and Y\ ,
Yi, . ■ . , Ym, both simple random samples of size n and m from their respective normal distributions.
Then under Ho (that is, assuming //(l is true) we have /ix = fiy or rewriting, fix - fiY — 0, so
T =
X-Y
X-Y - (flX - fly)
S p
Ji + 1
y n m
~ t(df = n + m - 2).
(10.4.3)
10.4.1 Independent Samples
Remark 10.14. If the values of crx and crY are known, then we can plug them in to our statistic:
X — Y
Z = — : (10.4.4)
Jcrx/n + CTy/m
the result will have a norm(mean = 0, sd = 1) distribution when Ho : fix = fiY is true.
Remark 10.15. Even if the values of crx and crY are not known, if both n and m are large then we
can plug in the sample estimates and the result will have approximately a norm(mean = 0, sd = 1)
distribution when Ho : fix — fiy is true.
v _ y
Z = — . (10.4.5)
■Jsx/n + Sy/m
Remark 10.16. It is usually important to construct side-by-side boxplots and other visual displays
in concert with the hypothesis test. This gives a visual comparison of the samples and helps to
identify departures from the test’s assumptions - such as outliers.
Remark 10.17. WATCH YOUR ASSUMPTIONS.
• The normality assumption can be relaxed as long as the population distributions are not
highly skewed.
• The equal variance assumption can be relaxed as long as both sample sizes n and m are large.
However, if one (or both) samples is small, then the test does not perform well; we should
instead use the methods of Chapter 13.
For a nonparametric alternative to the two-sample F test see Chapter 15.
10.4.2 Paired Samples
10.4.3 How to do it with R
> t . test (extra ~ group, data = sleep, paired = TRUE)
10.5. OTHER HYPOTHESIS TESTS
241
Paired t-test
data: extra by group
t = -4. 0621, df = 9, p-value = 0.002833
alternative hypothesis: true difference in means is not equal to 0
95 percent confidence interval:
-2.4598858 -0.7001142
sample estimates:
mean of the differences
-1.58
10.5 Other Hypothesis Tests
10.5.1 Kolmogorov-Smirnov Goodness-of-Fit Test
10.5.2 How to do it with R
> ks.test(randu$x, "punif")
One-sample Kolmogorov-Smirnov test
data: randuSx
D = 0.0555, p-value = 0.1697
alternative hypothesis: two-sided
10.5.3 Shapiro- Wilk Normality Test
10.5.4 How to do it with R
> shapiro. test (womenSheight)
Shapiro-Wilk normality test
data: womenSheight
W = 0.9636, p-value = 0.7545
10.6 Analysis of Variance
10.6.1 How to do it with R
I am thinking
> with(chickwts, byCweight, feed, shapiro. test))
242 CHAPTER 10. HYPOTHESIS TESTING
feed: casein
Shapiro-Wilk normality test
data: dd[x, ]
W = 0.9166, p-value =
= 0.2592
feed: horsebean
Shapiro-Wilk normality test
data: dd[x, ]
W = 0.9376, p-value =
= 0.5265
feed: linseed
Shapiro-Wilk normality test
data: dd[x, ]
W = 0.9693, p-value =
= 0.9035
feed: meatmeal
Shapiro-Wilk normality test
data: dd[x, ]
W = 0.9791, p-value =
= 0.9612
feed: soybean
Shapiro-Wilk normality test
data: dd[x, ]
W = 0.9464, p-value =
= 0.5064
feed: sunflower
Shapiro-Wilk normality test
data: dd[x, ]
W = 0.9281, p-value = 0.3603
1 0. 7. SAMPLE SIZE AND POWER
243
and
> temp <- lm(weight ~ feed, data = chickwts)
and
> anova(temp)
Analysis of Variance Table
Response: weight
Df Sum Sq Mean Sq F value Pr(>F)
feed 5 231129 46226 15.365 5.936e-l® ***
Residuals 65 195556 3®Q9
Signif. codes: 0 '***' Q.0Q1 '**' ®.®1 ®.®5 0.1 ' ' 1
Plot for the intuition of between versus within group variation.
Plots for the hypothesis tests:
10.7 Sample Size and Power
The power function of a test for a parameter 6 is
/3(9 ) = PfReject Hq), -oo < 9 < oo.
Here are some properties of power functions:
1. /3(9) < a for any 9 e ©o, and /3(9q) = a. We interpret this by saying that no matter what value
9 takes inside the null parameter space, there is never more than a chance of a of rejecting
the null hypothesis. We have controlled the Type I error rate to be no greater than a.
2. lim n^ccfiW) = 1 for any fixed 9 e ©i. In other words, as the sample size grows without
bound we are able to detect a nonnull value of 9 with increasing accuracy, no matter how
close it lies to the null parameter space. This may appear to be a good thing at first glance,
but it often turns out to be a curse. For another interpretation is that our Type II error rate
grows as the sample size increases.
10.7.1 How to do it with R
I am thinking about replicate here, and also power . examp from the TeachingDemos package.
There is an even better plot in upcoming work from the HH package.
244
CHAPTER 10. HYPOTHESIS TESTING
Index
Figure 10.6.1: Between group versus within group variation
1 0. 7. SAMPLE SIZE AND POWER
245
Histogram of y2
CD
2 3 4 5 6 7
y2
Figure 10.6.2: Between group versus within group variation
F density
246
CHAPTER 10. HYPOTHESIS TESTING
F density: v-| = 5 v2 = 30
3
4/
f 2.534 0.05
f 3
Figure 10.6.3: Some F plots from the HH package
1 0. 7. SAMPLE SIZE AND POWER
247
se = 1 .00 z* = 1 .64 power = 0.26
n= 1 sd = 1.00 diff = 1 .00 alpha = 0.050
Null Distribution
-2-1 0 1 ' 2 3 4
x
Alternative Distribution
-2-1 0 1 ' 2 3 4
x
Figure 10.7.1: Plot of significance level and power
This graph was generated by the power . examp function from the TeachingDemos package. The plot corre-
sponds to the hypothesis test H0 : p = p0 versus H\ : p = pt (where p0 = 0 and pi = 1, by default) based on
a single observation X ~ norm(mean = /j, sd = cr). The top graph is of the H0 density while the bottom is of
the H i density. The significance level is set at a = 0.05, the sample size is n — 1, and the standard deviation
is cr = 1. The pink area is the significance level, and the critical value a.05 ~ 1.645 is marked at the left
boundary - this defines the rejection region. When Ho is true, the probability of falling in the rejection region
is exactly a = 0.05. The same rejection region is marked on the bottom graph, and the probability of falling in
it (when is true) is the blue area shown at the top of the display to be approximately 0.26. This probability
represents the power to detect a non-null mean value of p = 1 . With the command the run . power . examp 0
at the command line the same plot opens, but in addition, there are sliders available that allow the user to inter-
actively change the sample size n, the standard deviation cr, the true difference between the means p\ -po, and
the significance level a. By playing around the student can investigate the effect each of the aforementioned
parameters has on the statistical power. Note that you need the tkrplot package for run . power . examp.
248
CHAPTER 10. HYPOTHESIS TESTING
Chapter Exercises
Chapter 11
Simple Linear Regression
What do I want them to know?
• basic philosophy of SLR and the regression assumptions
• point and interval estimation of the model parameters, and how to use it to make predictions
• point and interval estimation of future observations from the model
• regression diagnostics, including R 2 and basic residual analysis
• the concept of influential versus outlying observations, and how to tell the difference
11.1 Basic Philosophy
Here we have two variables X and Y. For our purposes, X is not random (so we will write x), but Y
is random. We believe that Y depends in some way on x. Some typical examples of (x, Y) pairs are
• x - study time and Y = score on a test.
• x - height and Y = weight.
• x = smoking frequency and Y = age of first heart attack.
Given information about the relationship between x and Y, we would like to predict future values
of Y for particular values of x. This turns out to be a difficult problem1, so instead we first tackle
an easier problem: we estimate E Y. How can we accomplish this? Well, we know that Y depends
somehow on x, so it stands to reason that
E Y — p(x), a function of x. (11.1.1)
But we should be able to say more than that. To focus our efforts we impose some structure on the
functional form of p. For instance,
• if /i(x) = /Jo + f3\x, we try to estimate Pn and P\ .
1 Yogi Berra once said, "It is always difficult to make predictions, especially about the future.”
249
250
CHAPTER 11. SIMPLE LINEAR REGRESSION
• if p(x) = /Jo + Pix + P2X2 , we try to estimate /?(), /3\, and #>.
• if p(x) = p0^lX, we try to estimate /?o and f>\.
This helps us in the sense that we concentrate on the estimation of just a few parameters, /j(l and
ySi, say, rather than some nebulous function. Our modus operand i is simply to perform the random
experiment n times and observe the n ordered pairs of data (x\ , Fi), (X2, F2), . . . , (x„, Y„). We use
these n data points to estimate the parameters.
More to the point, there are three simple linear regression (SLR) assumptions that will form the
basis for the rest of this chapter:
Assumption 11.1. We assume that p is a linear function of x, that is,
p(x)=Po+P\X, (11.1.2)
where /3q and f3\ are unknown constants to be estimated.
Assumption 11.2. We further assume that F, is p(Xj) - the “signal” - plus some “error” (repre-
sented by the symbol ef:
Yi = j80 + P\xi + et, i = 1, 2, . . . ,n. (11.1.3)
Assumption 11.3. We lastly assume that the errors are i.i.d. normal with mean 0 and variance cr2:
e\, 62 ,...,en~ norm(mean = 0, sd = cr). (11.1.4)
Remark 1 1 .4. We assume both the normality of the errors e and the linearity of the mean function
p. Recall from Proposition 7.27 of Chapter 7 that if (A, F) ~ mvnorm then the mean of Y\x is a
linear function of x. This is not a coincidence. In more advanced classes we study the case that
both X and F are random, and in particular, when they are jointly normally distributed.
What does it all mean?
See Figure 11.1.1. Shown in the figure is a solid line, the regression line p, which in this display
has slope 0.5 and v-intercept 2.5, that is, p(x) = 2.5 + 0.5x. The intuition is that for each given value
of jr, we observe a random value of F which is normally distributed with a mean equal to the height
of the regression line at that x value. Normal densities are superimposed on the plot to drive this
point home; in principle, the densities stand outside of the page, perpendicular to the plane of the
paper. The figure shows three such values of x, namely, x = 1, x = 2.5, and x = 4. Not only do we
assume that the observations at the three locations are independent, but we also assume that their
distributions have the same spread. In mathematical terms this means that the normal densities all
along the line have identical standard deviations - there is no “fanning out” or “scrunching in” of
the normal densities as x increases2.
Example 11.5. Speed and stopping distance of cars. We will use the data frame cars from the
datasets package. It has two variables: speed and dist. We can take a look at some of the
values in the data frame:
-In practical terms, this constant variance assumption is often violated, in that we often observe scatterplots that fan out
from the line as x gets large or small. We say under those circumstances that the data show heteroscedasticity. There are
methods to address it, but they fall outside the realm of SLR.
11.1. BASIC PHILOSOPHY
251
Figure 11.1.1: Philosophical foundations of SLR
dist
252
CHAPTER 11. SIMPLE LINEAR REGRESSION
speed
Figure 1 1.1.2: Scatterplot of dist versus speed for the cars data
> head(cars)
speed dist
1 4 2
2 4 1®
3 7 4
4 7 22
5 8 16
6 9 1®
The speed represents how fast the car was going (x) in miles per hour and dist (T) measures
how far it took the car to stop, in feet. We can make a simple scatterplot of the data with the
command plot (dist ~speed, data = cars).
You can see the output in Figure 1 1.1.2, which was produced by the following code.
11.2. ESTIMATION
253
> plot(dist ~ speed, data = cars )
There is a pronounced upward trend to the data points, and the pattern looks approximately
linear. There does not appear to be substantial fanning out of the points or extreme values.
11.2 Estimation
11.2.1 Point Estimates of the Parameters
Where is p(x)2 In essence, we would like to “fit” a line to the points. But how do we determine a
“good” line? Is there a best line? We will use maximum likelihood to find it. We know:
Yi - Pq +PiXj + €i, i=\,...,n, (11.2.1)
where the e,’s are i.i.d. norm(mean = 0, sd = cr). Thus T,- ~ norm(mean = /3q + p\ Xj, sd = cr), i =
1 Furthermore, Fi Yn are independent - but not identically distributed. The likelihood
function is:
L(/?0,/?i,cr) =f[ /*(*>■
1=1
= J J(27rcr2) 1/2 exp
i=i
-{yt -p0 - Pi XjY
2 cr2
=(2ncr2) '^2 exp
-ZtiCv/ ~A) -P\xd2
2cr2
We take the natural logarithm to get
lnL(p0,Pi,cr) = ~ ln(27rcr2) -
Y1U (yt -Po ~P\Xj)2
2 cr2
(11.2.2)
(11.2.3)
(11.2.4)
(11.2.5)
We would like to maximize this function of fio and fi\ . See Appendix E.6 which tells us that we
should find critical points by means of the partial derivatives. Let us start by differentiating with
respect to jd<y.
d 1 ^
— In L = 0 - 2 2(y«- - A - A *«)(- 1), d 1 .2.6)
and the partial derivative equals zero when Y!l= \ (}’i - A) ~ P\xd = 0, that is, when
nPo+J3i^ixi = ^jyi. (11.2.7)
;=i i= l
Moving on, we next take the partial derivative of In L (Equation 1 1.2.5) with respect to (3\ to get
dpi
1 n
lnL = ° _ Txr2 ^ 2('y‘ ~Pixi)(-Xi),
1 n
: ^2 Tj (X,y' ~ P°Xi
(11.2.8)
(11.2.9)
254
CHAPTER 11. SIMPLE LINEAR REGRESSION
and this equals zero when the last sum equals zero, that is, when
n n n
PoYjXi+PlYl^
1=1 1=1
= Zx'>'-
i=i
(11.2.10)
Solving the system of equations 11.2.7 and 11.2.10
n
nPo +P\ y] Xi
1=1
ii
br
(11.2.11)
n n
Po'Zxi+Pi'L*
1=1 i= 1
n
- Yjx>y*
i=l
(11.2.12)
for pQ and P\ (in Exercise 1 1.2) gives
^ _ ZU x<yi - (Z/U xi) (E”=i yi
and
Po=y~/3iX. (11.2.14)
The conclusion? To estimate the mean line
p(x) =p0 + P\X, (11.2.15)
we use the “line of best fit”
/K*)=j Bo+Pix, (11.2.16)
where j3o and /3\ are given as above. For notation we will usually write bo = Po and b\ = P\ so that
p(x) — bo + b\x.
Remark 1 1.6. The formula for bt in Equation 1 1.2.13 gets the job done but does not really make
any sense. There are many equivalent formulas for b\ that are more intuitive, or at the least are
easier to remember. One of the author’s favorites is
b\ = r—, (11.2.17)
s*
where r, sy, and sx are the sample correlation coefficient and the sample standard deviations of the
Y and x data, respectively. See Exercise 1 1.3. Also, notice the similarity between Equation 1 1.2.17
and Equation 7.6.7.
How to do it with R
Here we go. R will calculate the linear regression line with the lm function. We will store the result
in an object which we will call cars . lm. Here is how it works:
)/-
(11.2.13)
> cars.lm <- lm(dist ~ speed, data = cars )
11.2. ESTIMATION
255
speed
Figure 11.2.1: Scatterplot with added regression line for the cars data
The first part of the input to the lm function, dist~speed, is a model formula, read as “dist is
described by speed”. The data = cars argument tells R where to look for the variables quoted in
the model formula. The output object cars . lm contains a multitude of information. Let’s first take
a look at the coefficients of the fitted regression line, which are extracted by the coef function’:
> coef (cars . lm)
(Intercept) speed
-17.579095 3.932409
The parameter estimates bo and b\ for the intercept and slope, respectively, are shown above.
The regression line is thus given by //(speed) = -17.58 + 3.93speed.
It is good practice to visually inspect the data with the regression line added to the plot. To do
this we first scatterplot the original data and then follow with a call to the abline function. The
inputs to abline are the coefficients of cars . lm (see Figure 1 1.2.1):
3 Alternatively, we could just type cars . lm to see the same thing.
256
CHAPTER 11. SIMPLE LINEAR REGRESSION
> plot(dist ~ speed, data = cars, pch = 16)
> abline(coef(cars))
To calculate points on the regression line we may simply plug the desired x value(s) into p,
either by hand, or with the predict function. The inputs to predict are the fitted linear model
object, cars . lm, and the desired x value(s) represented by a data frame. See the example below.
Example 11.7. Using the regression line for the cars data:
1. What is the meaning of p( 60) = fto + yS[(8)?
This represents the average stopping distance (in feet) for a car going 8 mph.
2. Interpret the slope .
The true slope f)\ represents the increase in average stopping distance for each mile per
hour faster that the car drives. In this case, we estimate the car to take approximately 3.93
additional feet to stop for each additional mph increase in speed.
3. Interpret the intercept ySo.
This would represent the mean stopping distance for a car traveling 0 mph (which our regres-
sion line estimates to be -17.58). Of course, this interpretation does not make any sense for
this example, because a car travelling Omph takes Oft to stop (it was not moving in the first
place)! What went wrong? Looking at the data, we notice that the smallest speed for which
we have measured data is 4 mph. Therefore, if we predict what would happen for slower
speeds then we would be extrapolating, a dangerous practice which often gives nonsensical
results.
11.2.2 Point Estimates of the Regression Line
We said at the beginning of the chapter that our goal was to estimate p = HY, and the arguments
in Section 11.2.1 showed how to obtain an estimate p of p when the regression assumptions hold.
Now we will reap the benefits of our work in more ways than we previously disclosed. Given a
particular value xo, there are two values we would like to estimate:
1 . the mean value of Y at xo, and
2. a future value of Y at xo-
The first is a number, p(x o), and the second is a random variable, T(xo), but our point estimate is
the same for both: p(x o).
Example 11.8. We may use the regression line to obtain a point estimate of the mean stopping
distance for a car traveling 8 mph: p( 15) - bo + 8/?i « -17.58 +(8) (3.93)» 13.88. We would also
use 13.88 as a point estimate for the stopping distance of a future car traveling 8 mph.
Note that we actually have observed data for a car traveling 8 mph; its stopping distance was
16 ft as listed in the fifth row of the cars data:
> cars [5, ]
11.2. ESTIMATION
257
speed dist
5 8 16
There is a special name for estimates j)(x o) when xq matches an observed value x; from the data
set. They are called fitted values, they are denoted by Y\, Y2, ..., Y„ (ignoring repetition), and they
play an important role in the sections that follow.
In an abuse of notation we will sometimes write Y or T(xo) to denote a point on the regression
line even when xo does not belong to the original data if the context of the statement obviates any
danger of confusion.
We saw in Example 11.7 that spooky things can happen when we are cavalier about point
estimation. While it is usually acceptable to predict/estimate at values of Xo that fall within the
range of the original x data, it is reckless to use jj for point estimates at locations outside that range.
Such estimates are usually worthless. Do not extrapolate unless there are compelling external
reasons, and even then, temper it with a good deal of caution.
How to do it with R
The fitted values are automatically computed as a byproduct of the model fitting procedure and
are already stored as a component of the cars . lm object. We may access them with the fitted
function (we only show the first five entries):
> fitted(cars . lm) [1 : 5]
1 2 3 4 5
-1 . 84946® -1.849460 9.947766 9.947766 13.880175
Predictions at x values that are not necessarily part of the original data are done with the
predict function. The first argument is the original cars. Ira object and the second argument
newdata accepts a dataframe (in the same form that was used to fit cars . lm) that contains the
locations at which we are seeking predictions.
Let us predict the average stopping distances of cars traveling 6 mph, 8 mph, and 21 mph:
> predict (cars . lm, newdata = data, frame (speed = c(6, 8, 21)))
1 2 3
6.015358 13.880175 65.001489
Note that there were no observed cars that traveled 6 mph or 21 mph. Also note that our estimate
for a car traveling 8 mph matches the value we computed by hand in Example 11.8.
11.2.3 Mean Square Error and Standard Error
To find the MLE of cr 2 we consider the partial derivative
(1L218)
and after plugging in fio and and setting equal to zero we get
o-2 = - V(y,- -Po — Pi x(')2 = - V [y, - jXx,)]2.
n n
1=1
(11.2.19)
258
CHAPTER 11. SIMPLE LINEAR REGRESSION
We write Yi = (i(Xi), and we let E, = Y , - K, be the fi1 residual. We see
n
ncr2 = ^ E2 = SS E = the sum of squared errors. (11.2.20)
i= t
For a point estimate of <x2 we use the mean square error S 2 defined by
= (11.2.21)
n — 2
and we estimate cr with the standard error S = Vs2. 4
How to do it with R
The residuals for the model may be obtained with the residuals function; we only show the first
few entries in the interest of space:
> residuals (cars . lm) [1 : 5]
1 2 3 4 5
3.84946® 11.84946® -5.947766 12.S52234 2.119825
In the last section, we calculated the fitted value for x — 8 and found it to be approximately
//( 8) 13.88. Now, it turns out that there was only one recorded observation at x = 8, and we have
seen this value in the output of head(cars) in Example 11.5; it was dist = 16 ft for a car with
speed = 8 mph. Therefore, the residual should be E = Y - Y which is E & 16-13.88. Now take a
look at the last entry of residuals(cars . lm), above. It is not a coincidence.
The estimate S for cr is called the Residual standard error and for the cars data is shown
a few lines up on the summary (cars . lm) output (see How to do it with R in Section 11.2.4). We
may read it from there to be S ~ 15.38, or we can access it directly from the summary object.
> carsumry <- summary (cars . lm)
> carsumry$sigma
[1] 15.37959
11.2.4 Interval Estimates of the Parameters
We discussed general interval estimation in Chapter 9. There we found that we could use what we
know about the sampling distribution of certain statistics to construct confidence intervals for the
parameter being estimated. We will continue in that vein, and to get started we will determine the
sampling distributions of the parameter estimates, b\ and /;(l.
To that end, we can see from Equation 11.2.13 (and it is made clear in Chapter 12) that b\ is
just a linear combination of normally distributed random variables, so b\ is normally distributed
too. Further, it can be shown that
b\ ~ norm (mean = f}\, sd = crbl) (11.2.22)
4Be careful not to confuse the mean square error S 2 with the sample variance S 2 in Chapter 3. Other notation the reader
may encounter is the lowercase s2 or the bulky MS E.
11.2. ESTIMATION
259
where
Q~b\
CT
(11.2.23)
is called the standard error of b\ which unfortunately depends on the unknown value of cr. We do
not lose heart, though, because we can estimate cr with the standard error S from the last section.
This gives us an estimate S for ab] defined by
S
*i
S
Veluz-x)2'
(11.2.24)
Now, it turns out that bo, b\ , and S are mutually independent (see the footnote in Section 12.2.7).
Therefore, the quantity
T = 1 (11.2.25)
5 b\
has a t(df = n — 2) distribution. Therefore, a 100(1 - a)% confidence interval for fi\ is given by
b i ±tff/2(df = 71-1)5*;
(11.2.26)
It is also sometimes of interest to construct a confidence interval for /3o in which case we will
need the sampling distribution of bo. It is shown in Chapter 12 that
b0 ~ norm (mean = A), sd = crbo) ,
(11.2.27)
where crbo is given by
1 x2
°To = - +
n 2"=i (xi-x)2’
and which we estimate with the S bo defined by
(11.2.28)
c o 1 ^
S hr. ~ S A I H
Thus the quantity
T =
n H"=i (Xi-x)1'
bo - A)
(11.2.29)
(11.2.30)
has a t(df — n - 2) distribution and a 100(1 - a)% confidence interval for A) is given by
b0±ta/2(df = n-l)Sbo (11.2.31)
How to do it with R
Let us take a look at the output from summary (cars . lm):
> summary (car s.lm)
260
CHAPTER 11. SIMPLE LINEAR REGRESSION
Call:
lm(formula = dist ~ speed, data = cars)
Residuals:
Min IQ Median 3Q Max
-29.069 -9.525 -2.272 9.215 43.201
Coefficients :
Estimate Std. Error t value Pr(>|t|)
(Intercept) -17.5791 6.7584 -2.601 0.0123
speed 3.9324 0.4155 9.464 1.49e-12
Signif. codes: 0
0.001
0.01 0.05 ' . ' 0.1 ' ' 1
Residual standard error: 15.38 on 48 degrees of freedom
Multiple R-squared: 0.6511, Adjusted R-squared: 0.6438
F-statistic: 89.57 on 1 and 48 DF, p-value: 1.490e-12
In the Coefficients section we find the parameter estimates and their respective standard
errors in the second and third columns; the other columns are discussed in Section 11.3. If we
wanted, say, a 95% confidence interval for fl\ we could use h\ = 3.932 and .S'/,, = 0.416 together
with a to.o25(df = 23) critical value to calculate b\ + to.o25(df = 23)5 ^ .
Or, we could use the confint function.
> confint (car s.lm)
2.5 % 97.5 %
(Intercept) -31.167850 -3.990340
speed 3.096964 4.767853
With 95% confidence, the random interval [3.097, 4.768] covers the parameter f}\.
11.2.5 Interval Estimates of the Regression Line
We have seen how to estimate the coefficients of regression line with both point estimates and
confidence intervals. We even saw how to estimate a value p{ x) on the regression line for a given
value of x, such as x = 15.
But how good is our estimate /)( 1 5)? How much confidence do we have in this estimate?
Furthermore, suppose we were going to observe another value of Y at x = 15. What could we say?
Intuitively, it should be easier to get bounds on the mean (average) value of Y at xq (called a
confidence interval for the mean value ofY at xq) than it is to get bounds on a future observation
of Y (called a prediction interval for Y at x(t). As we shall see, the intuition serves us well and
confidence intervals are shorter for the mean value, longer for the individual value.
Our point estimate of p(xo) is of course Y = F(xq), so for a confidence interval we will need to
know F’s sampling distribution. It turns out (see Section ) that Y = jj(xo) is distributed
Y ~ norm
(11.2.32)
11.2. ESTIMATION
261
Since u is unknown we estimate it with S (we should expect the appearance of a t(df = n — 2)
distribution in the near future).
A 100(1 - a)% confidence interval (Cl) for /j(xq ) is given by
f±w“ = " +
y\2
(11.2.33)
It is time for prediction intervals, which are slightly different. In order to find confidence bounds
for a new observation of Y (we will denote it Tnew) we use the fact that
knew ~ norm
, , , , I (XQ~X)2
mean = p(xo ), sd = cr^ 1 + - + — -
n 2"=i (xi-x)2
(11.2.34)
Of course cr is unknown and we estimate it with S . Thus, a 100(1 - a)% prediction interval (PI)
for a future value of Y at xq is given by
/ 1 (xn — X)2
Y(x0) ± ta/2(df = n-l)S Jl + - + ^~7 4^. (11.2.35)
V n i= i vT x)
We notice that the prediction interval in Equation 11.2.35 is wider than the confidence interval in
Equation 1 1.2.33, as we expected at the beginning of the section.
How to do it with R
Confidence and prediction intervals are calculated in R with the predict function, which we
encountered in Section 11.2.2. There we neglected to take advantage of its additional interval
argument. The general syntax follows.
Example 11.9. We will find confidence and prediction intervals for the stopping distance of a car
travelling 5, 6, and 21 mph (note from the graph that there are no collected data for these speeds).
We have computed cars . lm earlier, and we will use this for input to the predict function. Also,
we need to tell R the values of xo at which we want the predictions made, and store the xo values
in a data frame whose variable is labeled with the correct name. This is important.
> new <- data. frame (speed = c (5, 6, 21))
Next we instruct R to calculate the intervals. Confidence intervals are given by
> predict (cars . lm, newdata = new, interval = "confidence")
fit lwr upr
1 2.Q82949 -7.64415® 11.81805
2 6.015358 -2.973341 15.00406
3 65.001489 58.597384 71.40559
Prediction intervals are given by
> predict (cars . lm, newdata = new, interval = "prediction")
262
CHAPTER 11. SIMPLE LINEAR REGRESSION
fit
1 2.082949
2 6.015358
3 65.001489
lwr upr
-30.33359 34.49948
-26.18731 38.21803
33.42257 96.58040
The type of interval is dictated by the interval argument (which is none by default), and the
default confidence level is 95% (which can be changed with the level argument).
Example 11.10. Using the cars data,
1 . Report a point estimate of and a 95% confidence interval for the mean stopping distance for
a car travelling 5 mph.
The fitted value for x = 5 is 2.08, so a point estimate would be 2.08 ft. The 95% Cl is
given by [-7.64, 1 1.81], so with 95% confidence the mean stopping distance lies somewhere
between -7.64 ft and 11.81 ft.
2. Report a point prediction for and a 95% prediction interval for the stopping distance of a
hypothetical car travelling 21 mph.
The fitted value for x = 21 is 65, so a point prediction for the stopping distance is 65 ft.
The 95% PI is given by [33.42, 96.58], so with 95% confidence we may assert that the
hypothetical stopping distance for a car travelling 21 mph would lie somewhere between
33.42 ft and 96.58 ft.
Graphing the Confidence and Prediction Bands
We earlier guessed that a bound on the value of a single new observation would be inherently less
certain than a bound for an average (mean) value; therefore, we expect the CIs for the mean to be
tighter than the Pis for a new observation. A close look at the standard deviations in Equations
11.2.33 and 1 1.2.35 confirms our guess, but we would like to see a picture to drive the point home.
We may plot the confidence and prediction intervals with one fell swoop using the ci . plot
function from the HH package [40]. The graph is displayed in Figure 1 1.2.2.
> library(HH)
> ci .plot (cars . lm)
Notice that the bands curve outward away from the regression line as the x values move away
from the center. This is expected once we notice the (xo - x)2 term in the standard deviation
formulas in Equations 11.2.33 and 11.2.35.
11.3 Model Utility and Inference
11.3.1 Hypothesis Tests for the Parameters
Much of the attention of SLR is directed toward j3i because when /? i f 0 the mean value of Y
increases (or decreases) as x increases. Further, if =0 then the mean value of Y remains the
same, regardless of the value of x (when the regression assumptions hold, of course). It is thus very
dist
11.3. MODEL UTILITY AND INFERENCE
263
100 -
50 -
0 -
95% confidence and prediction intervals for cars.lm
speed
observed •
fit
conf int
pred int
Figure 11.2.2: Scatterplot with confidence/prediction bands for the cars data
264
CHAPTER 11. SIMPLE LINEAR REGRESSION
important to decide whether or not (5 \ = 0. We address the question with a statistical test of the
null hypothesis Ho : j$\ — 0 versus the alternative hypothesis 11 \ : fl\ 4 0, and to do that we need to
know the sampling distribution of b\ when the null hypothesis is true.
To this end we already know from Section 1 1.2.4 that the quantity
b i -jSi
S b,
(11.3.1)
has a t(df = n - 2) distribution; therefore, when [f - 0 the quantity b\ /Si,, has a t(df = n - 2)
distribution and we can compute a p- value by comparing the observed value of h\ /S with values
under a t(df = n - 2) curve.
Similarly, we may test the hypothesis Hq : j3o = 0 versus the alternative II \ : /3q 4 0 with the
statistic T = bo/S b0, where S is given in Section 11.2.4. The test is conducted the same way as
for Pi.
How to do it with R
Let us take another look at the output from summary (cars . lm):
> summary (cars . lm)
Call:
lm(formula = dist ~ speed,
data
= cars)
Residuals:
Min IQ Median
3Q
Max
-29.869 -9.525 -2.272
9.215
43.281
Coefficients :
Estimate Std.
Error
t value
Pr(> | t | )
(Intercept) -17.5791 6
.7584
-2.681
8.8123
speed 3.9324 8
.4155
9.464
1.49e-12
Signif. codes: 8 '***' 8.
881 ' •
' 8.81
8.85
Residual standard error: 15.38 on 48 degrees of freedom
Multiple R-squared: 8.6511, Adjusted R-squared: 8.6438
F-statistic: 89.57 on 1 and 48 DF, p-value: 1.498e-12
In the Coefficients section we find the t statistics and the p- values associated with the tests
that the respective parameters are zero in the fourth and fifth columns. Since the p- values are
(much) less than 0.05, we conclude that there is strong evidence that the parameters /3\ 4 0 and
/3o 4 0, and as such, we say that there is a statistically significant linear relationship between dist
and speed.
11.3.2 Simple Coefficient of Determination
It would be nice to have a single number that indicates how well our linear regression model is
doing, and the simple coefficient of determination is designed for that purpose. In what follows,
11.3. MODEL UTILITY AND INFERENCE
265
we observe the values Y\ , Y2, ... ,T„, and the goal is to estimate p(xo), the mean value of Y at the
location xq.
If we disregard the dependence of Y and x and base our estimate only on the Y values then a
reasonable choice for an estimator is just the MLE of / 1 , which is Y . Then the errors incurred by the
estimate are just Y, - Y and the variation about the estimate as measured by the sample variance is
proportional to
n
SSTO = Y}Yi -Y)2- (11.3.2)
i=i
Here, SSTO is an acronym for the total sum of squares.
But we do have additional information, namely, we have values x, associated with each value
of Yj. We have seen that this information leads us to the estimate Y, and the errors incurred are just
the residuals, E, = Y, - Y,. The variation associated with these errors can be measured with
n
ssE = YjJi-Yd2. (n.3.3)
i= 1
We have seen the SSE before, which stands for the sum of squared errors or error sum of squares.
Of course, we would expect the error to be less in the latter case, since we have used more informa-
tion. The improvement in our estimation as a result of the linear regression model can be measured
with the difference
(Y, - Y) - (Y, - Yd = Yj - Y,
and we measure the variation in these errors with
n
SSR = YJ(Yi-Y)2, (11-3.4)
1=1
also known as the regression sum of squares. It is not obvious, but some algebra proved a famous
result known as the ANOVA Equality:
n n n
YjY, - Y)2 = Yj(Y- - F)2 + Z(y'- - f‘)2 (1 L3-5)
i— 1 i— 1 i— 1
or in other words.
SSTO = SSR + SSE.
(11.3.6)
This equality has a nice interpretation. Consider SS TO to be the total variation of the errors. Think
of a decomposition of the total variation into pieces: one piece measuring the reduction of error
from using the linear regression model, or explained variation ( S S R ), while the other represents
what is left over, that is, the errors that the linear regression model doesn’t explain, or unexplained
variation ( S S E). In this way we see that the ANOVA equality merely partitions the variation into
total variation = explained variation + unexplained variation.
For a single number to summarize how well our model is doing we use the simple coefficient of
determination r2, defined by
= 1 -
SSE
SSTO'
(11.3.7)
266
CHAPTER 11. SIMPLE LINEAR REGRESSION
We interpret r2 as the proportion of total variation that is explained by the simple linear regression
model. When r2 is large, the model is doing a good job; when r2 is small, the model is not doing a
good job.
Related to the simple coefficient of determination is the sample correlation coefficient, r. As
you can guess, the way we get r is by the formula |r[ = vr2. But how do we get the sign? It is
equal the sign of the slope estimate b\ . That is, if the regression line (tlx) has positive slope, then
r = yr2. Likewise, if the slope of (j(x) is negative, then r = - yr2.
How to do it with R
The primary method to display partitioned sums of squared errors is with an ANOVA table. The
command in R to produce such a table is anova. The input to anova is the result of an lm call
which for the cars data is cars . lm.
> anova (cars . lm)
Analysis of Variance Table
Response: dist
Df Sum Sq Mean Sq F value Pr(>F)
speed 1 21186 21185.5 89.567 1.490e-12 ***
Residuals 48 11354 236.5
Signif. codes: 0 '***' 0.001 '**' 0.01 0.05 0.1 ' ' 1
The output gives
r2 = 1 ^- = 1 11353-5 » 0.65.
SSR + SSE 21185.5 + 11353.5
The interpretation should be: “The linear regression line accounts for approximately 65% of the
variation of dist as explained by speed”.
The value of r2 is stored in the r. squared component of summary (cars . lm), which we
called carsumry.
> carsumrySr . squared
[1] 0.6510794
We already knew this. We saw it in the next to the last line of the summary(cars . lm) output
where it was called “Multiple R-squared". Listed right beside it is the Adjusted R-squared
which we will discuss in Chapter 12.
For the cars data, we find r to be
> sqrt (carsumrySr . squared)
[1] 0.8068949
We choose the principal square root because the slope of the regression line is positive.
11.4. RESIDUAL ANALYSIS
267
11.3.3 Overall F statistic
There is another way to test the significance of the linear regression model. In SLR, the new way
also tests the hypothesis //q : fi\ - 0 versus II \ [i\ 4- 0, but it is done with a new test statistic
called the overall F statistic. It is defined by
SSR
SSE/Oi- 2)'
(11.3.8)
Under the regression assumptions and when Hq is true, the F statistic has an f(dfl = 1, df2 =
n - 2) distribution. We reject H0 when F is large - that is, when the explained variation is large
relative to the unexplained variation.
All this being said, we have not yet gained much from the overall F statistic because we already
knew from Section 11.3.1 how to test Hy : fl\ = 0. . . we use the Student’s t statistic. What is worse
is that (in the simple linear regression model) it can be proved that the F in Equation 11.3.8 is
exactly the Student’s 1 statistic for /jj squared.
F =
(11.3.9)
So why bother to define the F statistic? Why not just square the t statistic and be done with it? The
answer is that the F statistic has a more complicated interpretation and plays a more important role
in the multiple linear regression model which we will study in Chapter 12. See Section 12.3.3 for
details.
11.3.4 How to do it with R
The overall F statistic and p- value are displayed in the bottom line of the summary (cars . lm)
output. It is also shown in the final columns of anovafcars . lm) :
> anovaCcars . lm)
Analysis of Variance Table
Response: dist
Df Sum Sq Mean Sq F value Pr(>F)
speed 1 21186 21185.5 89.567 1.490e-12 ***
Residuals 48 11354 236.5
Signif. codes: 0 '***' 0.001 '**' 0.01 0.05 0.1 ' ' 1
Here we see that the F statistic is 89.57 with a p- value very close to zero. The conclusion:
there is very strong evidence that Hq : ft i = 0 is false, that is, there is strong evidence that fi\ + 0.
Moreover, we conclude that the regression relationship between dist and speed is significant.
Note that the value of the F statistic is the same as the Student’s t statistic for speed squared.
11.4 Residual Analysis
We know from our model that Y = p(x) + e, or in other words, e = Y — p(x). Further, we know that
e ~ norm(mean = 0, sd = cr). We may estimate e ,• with the residual E ,• = T, - 7,-, where Y, = fi(Xj).
268
CHAPTER 11. SIMPLE LINEAR REGRESSION
If the regression assumptions hold, then the residuals should be normally distributed. We check
this in Section 1 1.4.1. Further, the residuals should have mean zero with constant variance <x2, and
we check this in Section 11.4.2. Last, the residuals should be independent, and we check this in
Section 1 1.4.3.
In every case, we will begin by looking at residual plots - that is, scatterplots of the residuals
Ej versus index or predicted values K, - and follow up with hypothesis tests.
11.4.1 Normality Assumption
We can assess the normality of the residuals with graphical methods and hypothesis tests. To check
graphically whether the residuals are normally distributed we may look at histograms or q-q plots.
We first examine a histogram in Figure 1 1 .4. 1 . There we see that the distribution of the residuals
appears to be mound shaped, for the most part. We can plot the order statistics of the sample
versus quantiles from a norm(mean = 0, sd = 1) distribution with the command plot (cars . lm,
which = 2), and the results are in Figure 11.4.1. If the assumption of normality were true, then
we would expect points randomly scattered about the dotted straight line displayed in the figure. In
this case, we see a slight departure from normality in that the dots show systematic clustering on
one side or the other of the line. The points on the upper end of the plot also appear begin to stray
from the line. We would say there is some evidence that the residuals are not perfectly normal.
Testing the Normality Assumption
Even though we may be concerned about the plots, we can use tests to determine if the evidence
present is statistically significant, or if it could have happened merely by chance. There are many
statistical tests of normality. We will use the Shapiro-Wilk test, since it is known to be a good
test and to be quite powerful. However, there are many other fine tests of normality including the
Anderson-Darling test and the Lillefors test, just to mention two of them.
The Shapiro-Wilk test is based on the statistic
where the are the ordered residuals and the a, are constants derived from the order statistics of
a sample of size n from a normal distribution. See Section 10.5.3.
We perform the Shapiro-Wilk test below, using the shapiro. test function from the stats
package. The hypotheses are
W =
(11.4.1)
//o : the residuals are normally distributed
versus
H i : the residuals are not normally distributed.
The results from R are
> shapiro. test(residuals(cars . lm))
Standardized residuals
11.4. RESIDUAL ANALYSIS
269
Normal Q-Q
-2-10 1 2
Theoretical Quantiles
lm(dist ~ speed)
Figure 11.4.1: Normal q-q plot of the residuals for the cars data
Used for checking the normality assumption. Look out for any curvature or substantial departures from the
straight line; hopefully the dots hug the line closely.
270
CHAPTER 11. SIMPLE LINEAR REGRESSION
Shapiro-Wilk normality test
data: residuals (cars . lm)
W = 0.9451, p-value = 0.02153
For these data we would reject the assumption of normality of the residuals at the a = 0.05
significance level, but do not lose heart, because the regression model is reasonably robust to de-
partures from the normality assumption. As long as the residual distribution is not highly skewed,
then the regression estimators will perform reasonably well. Moreover, departures from constant
variance and independence will sometimes affect the quantile plots and histograms, therefore it is
wise to delay final decisions regarding normality until all diagnostic measures have been investi-
gated.
11.4.2 Constant Variance Assumption
We will again go to residual plots to try and determine if the spread of the residuals is changing
over time (or index). However, it is unfortunately not that easy because the residuals do not have
constant variance! In fact, it can be shown that the variance of the residual £j- is
Var(£j) = a2{ 1 - hu), i = 1, 2, . . . ,n, (1 1.4.2)
where ha is a quantity called the leverage which is defined below. Consequently, in order to check
the constant variance assumption we must standardize the residuals before plotting. We estimate the
standard error of E, with se, = s VO _ hu) and define the standardized residuals Rt, i = 1,2
by
Ri = 7= , i= 1,2, ...,n. (11.4.3)
s VI - K
For the constant variance assumption we do not need the sign of the residual so we will plot
V|A; versus the fitted values. As we look at a scatterplot of VW versus Y, we would expect
under the regression assumptions to see a constant band of observations, indicating no change in
the magnitude of the observed distance from the line. We want to watch out for a fanning-out of
the residuals, or a less common funneling-in of the residuals. Both patterns indicate a change in the
residual variance and a consequent departure from the regression assumptions, the first an increase,
the second a decrease.
In this case, we plot the standardized residuals versus the fitted values. The graph may be seen
in Figure 11.4.2. For these data there does appear to be somewhat of a slight fanning-out of the
residuals.
Testing the Constant Variance Assumption
We will use the Breusch-Pagan test to decide whether the variance of the residuals is nonconstant.
The null hypothesis is that the variance is the same for all observations, and the alternative hypoth-
esis is that the variance is not the same for all observations. The test statistic is found by fitting a
linear model to the centered squared residuals
Wi = Ef -
SSE
i = 1, 2, .... n.
n
(11.4.4)
^Standardized residualsl
11.4. RESIDUAL ANALYSIS
271
Scale-Location
Fitted values
lm(dist ~ speed)
Figure 1 1.4.2: Plot of standardized residuals against the fitted values for the cars data
Used for checking the constant variance assumption. Watch out for any fanning out (or in) of the dots;
hopefully they fall in a constant band.
272
CHAPTER 11. SIMPLE LINEAR REGRESSION
By default the same explanatory variables are used in the new model which produces fitted values
W„ i = 1,2,..., n. The Breusch-Pagan test statistic in R is then calculated with
n n
BP = nYjWf + YjW?- (11.4.5)
i= l ;= l
We reject the null hypothesis if BP is too large, which happens when the explained variation in the
new model is large relative to the unexplained variation in the original model.
We do it in R with the bptest function from the lmtest package [93].
> library (lmtest)
> bptest (cars . lm)
studentized Breusch-Pagan test
data: cars. Ira
BP = 3.2149, df = 1, p-value = 8.07297
For these data we would not reject the null hypothesis at the a = 0.05 level. There is relatively
weak evidence against the assumption of constant variance.
11.4.3 Independence Assumption
One of the strongest of the regression assumptions is the one regarding independence. Departures
from the independence assumption are often exhibited by correlation (or autocorrelation, literally,
self-correlation) present in the residuals. There can be positive or negative correlation.
Positive correlation is displayed by positive residuals followed by positive residuals, and neg-
ative residuals followed by negative residuals. Looking from left to right, this is exhibited by a
cyclical feature in the residual plots, with long sequences of positive residuals being followed by
long sequences of negative ones.
On the other hand, negative correlation implies positive residuals followed by negative residu-
als, which are then followed by positive residuals, etc. Consequently, negatively correlated resid-
uals are often associated with an alternating pattern in the residual plots. We examine the residual
plot in Figure 1 1.4.3. There is no obvious cyclical wave pattern or structure to the residual plot.
Testing the Independence Assumption
We may statistically test whether there is evidence of autocorrelation in the residuals with the
Durbin- Watson test. The test is based on the statistic
D =
SUC E, - £,_!)2
ZjLt Ei
(11.4.6)
Exact critical values are difficult to obtain, but R will calculate the p-value to great accuracy. It is
performed with the dwtest function from the lmtest package. We will conduct a two sided test
that the correlation is not zero, which is not the default (the default is to test that the autocorrelation
is positive).
Residuals
11.4. RESIDUAL ANALYSIS
273
Residuals vs Fitted
Fitted values
lm(dist ~ speed)
Figure 1 1.4.3: Plot of the residuals versus the fitted values for the cars data
Used for checking the independence assumption. Watch out for any patterns or structure; hopefully the points
are randomly scattered on the plot.
274
CHAPTER 11. SIMPLE LINEAR REGRESSION
> library (lmtest)
> dwtest(cars . lm, alternative = "two. sided")
Durbin-Watson test
data: cars.lm
DW = 1.6762, p-value = 0.19Q4
alternative hypothesis: true autocorelation is not ®
In this case we do not reject the null hypothesis at the a = 0.05 significance level; there is very
little evidence of nonzero autocorrelation in the residuals.
11.4.4 Remedial Measures
We often find problems with our model that suggest that at least one of the three regression as-
sumptions is violated. What do we do then? There are many measures at the statistician’s disposal,
and we mention specific steps one can take to improve the model under certain types of violation.
Mean response is not linear. We can directly modify the model to better approximate the mean
response. In particular, perhaps a polynomial regression function of the form
p(x) — /?o +P\x\ +132x1
would be appropriate. Alternatively, we could have a function of the form
p(x ) =/3oeP'x.
Models like these are studied in nonlinear regression courses.
Error variance is not constant. Sometimes a transformation of the dependent variable will take
care of the problem. There is a large class of them called Box-Cox transformations. They
take the form
Y* = Ya, (11.4.7)
where A is a constant. (The method proposed by Box and Cox will determine a suitable value
of A automatically by maximum likelihood). The class contains the transformations
A = 2, Y* = Y2
A = 0.5, Y* = Vk
A = 0, r = In Y
a — — i, r = i/y
Alternatively, we can use the method of weighted least squares. This is studied in more detail
in later classes.
Error distribution is not normal. The same transformations for stabilizing the variance are equally
appropriate for smoothing the residuals to a more Gaussian form. In fact, often we will kill
two birds with one stone.
Errors are not independent. There is a large class of autoregressive models to be used in this
situation which occupy the latter part of Chapter 16.
11.5. OTHER DIAGNOSTIC TOOLS
275
11.5 Other Diagnostic Tools
There are two types of observations with which we must be especially careful:
Influential observations are those that have a substantial effect on our estimates, predictions, or
inferences. A small change in an influential observation is followed by a large change in the
parameter estimates or inferences.
Outlying observations are those that fall fall far from the rest of the data. They may be indicating
a lack of fit for our regression model, or they may just be a mistake or typographical error
that should be corrected. Regardless, special attention should be given to these observations.
An outlying observation may or may not be influential.
We will discuss outliers first because the notation builds sequentially in that order.
11.5.1 Outliers
There are three ways that an observation ( jc, ) may be an outlier: it can have an x,- value which
falls far from the other x values, it can have a yy value which falls far from the other y values, or it
can have both its x, and y, values falling far from the other x and y values.
Leverage
Leverage statistics are designed to identify observations which have x values that are far away from
the rest of the data. In the simple linear regression model the leverage of x,- is denoted by h,, and
defined by
1 (Xi - x)1
n + 2£=i(x*-x)2’
i = 1,2, ... ,n.
(11.5.1)
The formula has a nice interpretation in the SLR model: if the distance from x, to x is large relative
to the other x’s then h,, will be close to 1 .
Leverages have nice mathematical properties; for example, they satisfy
and their sum is
0 < ha < 1,
X h"
i=i
y [1 (Xj - x)2
t An + (**-*)2r
n £,-(x,- ~ x)2
n Tjflxk - xj2 ’
2.
(11.5.2)
(11.5.3)
(11.5.4)
(11.5.5)
A rule of thumb is to consider leverage values to be large if they are more than double their average
size (which is 2 /« according to Equation 1 1.5.5). So leverages larger than 4/n are suspect. Another
rule of thumb is to say that values bigger than 0.5 indicate high leverage, while values between 0.3
and 0.5 indicate moderate leverage.
276
CHAPTER 11. SIMPLE LINEAR REGRESSION
Standardized and Studentized Deleted Residuals
We have already encountered the standardized residuals r, in Section 1 1.4.2; they are merely resid-
uals that have been divided by their respective standard deviations:
Ri =
Ej
s vr^T
i — 1,2, ... . n.
(11.5.6)
Values of \Rt\ > 2 are extreme and suggest that the observation has an outlying v- value.
Now delete the 1th case and fit the regression function to the remaining n — 1 cases, producing a
fitted value Y(i) with deleted residual D , = F,- - Y&. It is shown in later classes that
Var(D,) =
1 ~hii
i — 1,2, ... ,n.
(11.5.7)
so that the studentized deleted residuals t, defined by
ti =
D;
S&Kl-huY
i = 1,2, ... ,n.
(11.5.8)
have a t( df = n - 3) distribution and we compare observed values of f,- to this distribution to decide
whether or not an observation is extreme.
The folklore in regression classes is that a test based on the statistic in Equation 1 1.5.8 can be
too liberal. A rule of thumb is if we suspect an observation to be an outlier before seeing the data
then we say it is significantly outlying if its two-tailed p- value is less than a, but if we suspect an
observation to be an outlier after seeing the data then we should only say it is significantly outlying
if its two-tailed /;- value is less than a/n. The latter rule of thumb is called the Bonferroni approach
and can be overly conservative for large data sets. The responsible statistician should look at the
data and use his/her best judgement, in every case.
11.5.2 How to do it with R
We can calculate the standardized residuals with the rstandard function. The input is the lm
object, which is cars . lm.
> sres <- rstandard(cars.lm)
> sres [1:5]
1 2 3 4 5
0 . 2660415 8.8189327 -8.4813462 8.8132663 8.1421624
We can find out which observations have studentized residuals larger than two with the com-
mand
> sres [which (abs (sres) > 2) ]
23 35 49
2.795166 2.827818 2.919868
In this case, we see that observations 23, 35, and 49 are potential outliers with respect to their
y- value.
We can compute the studentized deleted residuals with rstudent:
11.5. OTHER DIAGNOSTIC TOOLS
277
> sdelres <- rstudent(cars . lm)
> sdelres [1 : 5]
1 2 B 4 5
® . 263450® 0 .8160784 -0.3978115 0.8103526 0.1407033
We should compare these values with critical values from a t( df = n — 3) distribution, which in
this case is t(df = 50 - 3 = 47). We can calculate a 0.005 quantile and check with
> t0.®®5 <- qt(&. 6)6)5, df = 47, lower . tail = FALSE)
> sdelres [which(abs(sdelres) > t6).QlS5)]
23 49
3.022829 3.184993
This means that observations 23 and 49 have a large studentized deleted residual. The leverages
can be found with the hatvalues function:
> leverage <- hatvalues (cars . lm)
> leverage [1 : 5]
1 2 3 4 5
0.11486131 0.11486131 0.07150365 0.07150365 0.05997080
> leverage [which(leverage > 4/56))]
1 2 50
0.11486131 0.11486131 0.08727007
Here we see that observations 1, 2, and 50 have leverages bigger than double their mean value.
These observations would be considered outlying with respect to their x value (although they may
or may not be influential).
11.5.3 Influential Observations
DFBETAS and DFFITS
Anytime we do a statistical analysis, we are confronted with the variability of data. It is always
a concern when an observation plays too large a role in our regression model, and we would not
like or procedures to be overly influenced by the value of a single observation. Hence, it becomes
desirable to check to see how much our estimates and predictions would change if one of the
observations were not included in the analysis. If an observation changes the estimates/predictions
a large amount, then the observation is influential and should be subjected to a higher level of
scrutiny.
We measure the change in the parameter estimates as a result of deleting an observation with
DFBETAS . The DFBETAS for the intercept bo are given by
(DFBETAS) o(o = ~ b°i0 2 . i=l,2,...,n. (11.5.9)
s0 V » + Z"=ife-W
278
CHAPTER 11. SIMPLE LINEAR REGRESSION
and the DFBETAS for the slope b\ are given by
C DFBETAS)m= bl ~ bu° — i — 1,2, , n. (11.5.10)
s«) |2"=lte-x)2|
See Section 12.8 for a better way to write these. The signs of the DFBETAS indicate whether the
coefficients would increase or decrease as a result of including the observation. If the DFBETAS
are large, then the observation has a large impact on those regression coefficients. We label obser-
vations as suspicious if their DFBETAS have magnitude greater 1 for small data or 2 / yfn for large
data sets.
We can calculate the DFBETAS with the dfbetas function (some output has been omitted):
> dfb <- dfbetas (car s.lm)
> head(dfb)
(Intercept)
1 0. 89440188
2 0.29242487
3 -0.10749794
4 0.21897614
5 0.03407516
6 -0.11100703
speed
-0.08624563
-0.26715961
0.09369281
-0.19085472
-0.02901384
0.09174024
We see that the inclusion of the first observation slightly increases the Intercept and slightly
decreases the coefficient on speed.
We can measure the influence that an observation has on its fitted value with DFFIT S . These
are calculated by deleting an observation, refitting the model, recalculating the fit, then standardiz-
ing. The formula is
Y, - Yn)
(DFFITS)i = — — -S, i = 1,2, ... ,n. (11.5.11)
S (i) yfhn
The value represents the number of standard deviations of Y, that the fitted value Y, increases or
decreases with the inclusion of the 1th observation. We can compute them with the dffits function.
> dff <- dffits(cars . lm)
> dff[l : 5]
1 2 3 4 5
0.09490289 0.29397684 -0.11039550 0.22487854 0.03553887
A rule of thumb is to flag observations whose DFFIT exceeds one in absolute value, but there
are none of those in this data set.
Cook’s Distance
The DFFIT S are good for measuring the influence on a single fitted value, but we may want to
measure the influence an observation has on all of the fitted values simultaneously. The statistics
used for measuring this are Cook’s distances which may be calculated'’ by the formula
5Cook’s distances are actually defined by a different formula than the one shown. The formula in Equation 1 1.5.12 is
algebraically equivalent to the defining formula and is, in the author’s opinion, more transparent.
11.5. OTHER DIAGNOSTIC TOOLS
279
i = 1,2, ... ,n.
(11.5.12)
It shows that Cook’s distance depends both on the residual E ; and the leverage lia and in this way
Dj contains information about outlying x and y values.
To assess the significance of D, we compare to quantiles of an f(dfl = 2, df2 = n — 2)
distribution. A rule of thumb is to classify observations falling higher than the 50th percentile as
being extreme.
11.5.4 How to do it with R
We can calculate the Cook’s Distances with the cooks . distance function.
> cooksD <- cooks. distance (cars. lm)
> cooksD [1 : 5]
8.8845923121 8.8435139987 8.8862823583 8.8254673384 8.8886446785
We can look at a plot of the Cook’s distances with the command plot (cars . lm, which =
4).
Observations with the largest Cook’s D values are labeled, hence we see that observations 23,
39, and 49 are suspicious. However, we need to compare to the quantiles of an f(dfl = 2, df2 =
48) distribution:
> FIS. 56 <- qf(6. 5 , dfl = 2, df2 = 48)
> cooksD [which(coolcsD > F6.56 )]
named numeric (8)
We see that with this data set there are no observations with extreme Cook’s distance, after all.
11.5.5 All Influence Measures Simultaneously
We can display the result of diagnostic checking all at once in one table, with potentially influential
points displayed. We do it with the command influence . measures (car s . lm) :
> in fluence. measures (car s .lm)
The output is a huge matrix display, which we have omitted in the interest of brevity. A point
is identified if it is classified to be influential with respect to any of the diagnostic measures. Here
we see that observations 2, 1 1, 15, and 18 merit further investigation.
We can also look at all diagnostic plots at once with the commands
1
2
3
4
5
> par(mfrow = c(2, 2))
> plot (cars . lm)
> par(mfrow = c(l, 1))
Cook's distance
280
CHAPTER 11. SIMPLE LINEAR REGRESSION
Cook's distance
0 10 20 30 40 50
Obs. number
lm(dist ~ speed)
Figure 11.5.1: Cook’s distances for the cars data
Used for checking for influential and/our outlying observations. Values with large Cook’s distance merit
further investigation.
-^Standardized residualsl Residuals
11.5. OTHER DIAGNOSTIC TOOLS
281
Residuals vs Fitted
Normal Q-Q
Theoretical Quantiles
Scale-Location
Residuals vs Leverage
Figure 1 1.5.2: Diagnostic plots for the cars data
282
CHAPTER 11. SIMPLE LINEAR REGRESSION
The par command is used so that 2x2=4 plots will be shown on the same display. The
diagnostic plots for the cars data are shown in Figure 1 1.5.2:
We have discussed all of the plots except the last, which is possibly the most interesting. It
shows Residuals vs. Leverage, which will identify outlying y values versus outlying x values. Here
we see that observation 23 has a high residual, but low leverage, and it turns out that observations
1 and 2 have relatively high leverage but low/moderate leverage (they are on the right side of the
plot, just above the horizontal line). Observation 49 has a large residual with a comparatively large
leverage.
We can identify the observations with the identify command; it allows us to display the
observation number of dots on the plot. First, we plot the graph, then we call identify:
> plot (cars . lm, which = 59 # std'd resids vs lev plot
> identify (leverage , sres, n = 4) # identify 4 points
The graph with the identified points is omitted (but the plain plot is shown in the bottom right
corner of Figure 11.5.2). Observations 1 and 2 fall on the far right side of the plot, near the
horizontal axis.
11.5. OTHER DIAGNOSTIC TOOLS
283
Chapter Exercises
Exercise 11.1. Prove the ANOVA equality. Equation 1 1.3.5. Hint: show that
n
- YiXYi -Y) = 0.
;=i
Exercise 11.2. Solve the following system of equations for f3\ and /?o to find the MLEs for slope
and intercept in the simple linear regression model.
nPo+Pi ^Xi = ^
yi
Po Yj x< + A Y ^ = Y x‘y‘
i= 1 1=1 1=1
Exercise 11.3. Show that the formula given in Equation 1 1.2.17 is equivalent to
Pi =
2”= i xiyt - |
(e;-
=i x)
(2"
=i y<\
ln
22=1 x?
- 1
(27=i
Xif
ln
284
CHAPTER 11. SIMPLE LINEAR REGRESSION
Chapter 12
Multiple Linear Regression
We know a lot about simple linear regression models, and a next step is to study multiple regression
models that have more than one independent (explanatory) variable. In the discussion that follows
we will assume that we have p explanatory variables, where p > 1 .
The language is phrased in matrix terms - for two reasons. First, it is quicker to write and
(arguably) more pleasant to read. Second, the matrix approach will be required for later study of
the subject; the reader might as well be introduced to it now.
Most of the results are stated without proof or with only a cursory justification. Those yearning
for more should consult an advanced text in linear regression for details, such as Applied Linear
Regression Models [67]or Linear Models: Least Squares and Alternatives [69].
What do I want them to know?
• the basic MLR model, and how it relates to the SLR
• how to estimate the parameters and use those estimates to make predictions
• basic strategies to determine whether or not the model is doing a good job
• a few thoughts about selected applications of the MLR, such as polynomial, interaction, and
dummy variable models
• some of the uses of residuals to diagnose problems
• hints about what will be coming later
12.1 The Multiple Linear Regression Model
The first thing to do is get some better notation. We will write
>f
i
*11
*21 ' '
■ Xp\
Ynxl =
, and Xnx(p+i) —
i
*12
*22 •’
■ XP2
yn.
_i
X\n
n
Xpn-
285
286
CHAPTER 12. MULTIPLE LINEAR REGRESSION
The vector Y is called the response vector and the matrix X is called the model matrix. As in
Chapter 11, the most general assumption that relates Y to X is
Y = p(X) + e, (12.1.2)
where p is some function (the signal ) and e is the noise (everything else). We usually impose some
structure on p and e. In particular, the standard multiple linear regression model assumes
Y = X|3 + e, (12.1.3)
where the parameter vector (3 looks like
|3(P+i)xi = \Po . (12.1.4)
and the random vector enxi = |f i ei • • ■ e„ | is assumed to be distributed
e ~ mvnorm(mean = 0nxi, sigma = cr2InxnV (12.1.5)
The assumption on e is equivalent to the assumption that e\ , e 2 , ..., e„ are i.i.d. norm(mean =
0, sd = cr). It is a linear model because the quantity p(X) = X|3 is linear in the parameters /3(),
Pi,. . . , Pp. It may be helpful to see the model in expanded form; the above matrix formulation is
equivalent to the more lengthy
Yi = Po + PiXu + PiXn + ■ ■ ■ + PpXPi + 6i, i = 1,2, . . . ,n. (12.1.6)
Example 12.1. Girth, Height, and Volume for Black Cherry trees.Measurements were made
of the girth, height, and volume of timber in 31 felled black cherry trees. Note that girth is the
diameter of the tree (in inches) measured at 4 ft 6 in above the ground. The variables are
1. Girth: tree diameter in inches (denoted X\ )
2. Height: tree height in feet (*2).
3. Volume: volume of the tree in cubic feet, (y)
The data are in the datasets package and are already on the search path; they can be viewed with
> head(trees)
Girth Height
Volume
1
m
00
70
10.3
2
8.6
65
10.3
3
8.8
63
10.2
4
10.5
72
16.4
5
10.7
81
18.8
6
10.8
83
19.7
Let us take a look at a visual display of the data. For multiple variables, instead of a simple
scatterplot we use a scatterplot matrix which is made with the splom function in the lattice
package [75] as shown below. The plot is shown in Figure 12.1.1.
12.1. THE MULTIPLE LINEAR REGRESSION MODEL
287
o
°&s
#>°
<D
o
°°°o0 o
°00>^°
oo o
1 1 1 1
- 78 50 70
- 60
5<Volume4o -
30 -
10 30 20 -
i i i i 10-
6> °
1 1 1
“ 85 75 80 85
o
c
o
o
0
^ o %
- 80
do
6b00°
- 75Height75 -
« o°
o
o
o
o
70 -
o°o° °
o ° o
o °
65 70 75 65 _
i i i
o° o
o °
-20 1111
o
o
_ ^ 14161820
- 16
o O
CP
- u Girth u -
° °§°° CP
&§
12 -
8 101214 10 -
o009) <*6o
&
J 1 1 1 8_=
o
o
§
Scatter Plot Matrix
Figure 12.1.1: Scatterplot matrix of trees data
288
CHAPTER 12. MULTIPLE LINEAR REGRESSION
> library(lattice)
> splom(trees)
The dependent (response) variable Volume is listed in the first row of the scatterplot matrix.
Moving from left to right, we see an approximately linear relationship between Volume and the
independent (explanatory) variables Height and Girth. A first guess at a model for these data
might be
Y = Po + PiM + P2X2 + e, (12.1.7)
in which case the quantity p(x\ , xf) = fio + [3\X\ + P2X2 would represent the mean value of Y at the
point (t| , X2).
What does it mean?
The interpretation is simple. The intercept /3q represents the mean Volume when all other indepen-
dent variables are zero. The parameter /3,- represents the change in mean Volume when there is a
unit increase in x;, while the other independent variable is held constant. For the trees data, fi\
represents the change in average Volume as Girth increases by one unit when the Height is held
constant, and fh represents the change in average Volume as Height increases by one unit when
the Girth is held constant.
In simple linear regression, we had one independent variable and our linear regression surface
was ID, simply a line. In multiple regression there are many independent variables and so our
linear regression surface will be many-D. . . in general, a hyperplane. But when there are only
two explanatory variables the hyperplane is just an ordinary plane and we can look at it with a 3D
scatterplot.
One way to do this is with the R Commander in the Rcmdr package [31]. It has a 3D scatterplot
option under the Graphs menu. It is especially great because the resulting graph is dynamic; it can
be moved around with the mouse, zoomed, etc. But that particular display does not translate well
to a printed book.
Another way to do it is with the scatterplot3d function in the scatterplot3d package.
The code follows, and the result is shown in Figure 12.1.2.
> library (scatterplot3d)
> s3d <- with(trees , scatterplot3d(Girth, Height, Volume, pch = 16,
+ highlight . 3d = TRUE, angle = 6®))
> fit <- lm(Volume ~ Girth + Height, data = trees)
> s3d$plane3d(fit)
Looking at the graph we see that the data points fall close to a plane in three dimensional space.
(The plot looks remarkably good. In the author’s experience it is rare to see points fit so well to the
plane without some additional work.)
12.2 Estimation and Prediction
12.2.1 Parameter estimates
We will proceed exactly like we did in Section 11.2. We know
e ~ mvnormfmean = 0nxi, sigma = cr2Inxn), (12.2.1)
10 20 30 40 50 60 70 80
12.2. ESTIMATION AND PREDICTION
289
Girth
Figure 12.1.2: 3D scatterplot with regression plane for the trees data
Height
290
CHAPTER 12. MULTIPLE LINEAR REGRESSION
which means that Y = X|3 + e has an mvnorm (mean = X|3, sigma = cr2Inxn) distribution. There-
fore, the likelihood function is
L((3,fr)= 2^VeXP{'i(Y'XP)T(Y_X|3)}' (1Z2'2)
To maximize the likelihood in (3, we need to minimize the quantity g(|3) = (Y - X(3 )T (Y - X(3).
We do this by differentiating g with respect to |3. (It may be a good idea to brush up on the material
in Appendices E.5 and E.6.) First we will rewrite g :
g(t 3) = YtY - YtX|3 - |3tXtY + (3TXTX(3, (12.2.3)
which can be further simplified to g(|3) = YTY - 2[3TXTY + |3TXTX|3 since |3TXTY is 1 x 1 and
thus equal to its transpose. Now we differentiate to get
= 0 - 2XtY + 2XtX|3, (12.2.4)
3(3
since XTX is symmetric. Setting the derivative equal to the zero vector yields the so called “normal
equations”
XtX(3 = XtY. (12.2.5)
In the case that XTX is invertible1 , we may solve the equation for (3 to get the maximum likelihood
estimator of (3 which we denote by b:
b = (xTx)~‘ XtY. (12.2.6)
Remark 12.2. The formula in Equation 12.2.6 is convenient for mathematical study but is inconve-
nient for numerical computation. Researchers have devised much more efficient algorithms for the
actual calculation of the parameter estimates, and we do not explore them here.
Remark 12.3. We have only found a critical value, and have not actually shown that the critical
value is a minimum. We omit the details and refer the interested reader to [69].
12.2.2 How to do it with R
We do all of the above just as we would in simple linear regression. The powerhouse is the lm
function. Everything else is based on it. We separate explanatory variables in the model formula
by a plus sign.
> trees. lm <- lm(Volume ~ Girth + Height, data = trees)
> trees. lm
Call:
lm(formula = Volume ~ Girth + Height, data = trees)
Coefficients :
(Intercept) Girth Height
-57.9877 4. 7082 0.3393
1 We can find solutions of the normal equations even when XTX is not of full rank, but the topic falls outside the scope
of this book. The interested reader can consult an advanced text such as Rao [69].
12.2. ESTIMATION AND PREDICTION
291
We see from the output that for the trees data our parameter estimates are b = [-58.0 4.7 0.3],
and consequently our estimate of the mean response is p given by
p(x\ , x2) = b0 + bixi + b2X2, (12.2.7)
58.0 + 4.7xi +0.3x2. (12.2.8)
We could see the entire model matrix X with the model .matrix function, but in the interest of
brevity we only show the first few rows.
> head (model. matrix (trees. lm))
(Intercept) Girth Height
1 1 8.3 70
2 1 8.6 65
3 1 8.8 63
4 1 10.5 72
5 1 10.7 81
6 1 10.8 83
12.2.3 Point Estimates of the Regression Surface
The parameter estimates b make it easy to find the fitted values, Y. We write them individually as
Yi, i = 1 , 2, . . . , n, and recall that they are defined by
% = p(xu, x2i), (12.2.9)
= bo + b\X\j + b2x2i, i—l,2,...,n. (12.2.10)
They are expressed more compactly by the matrix equation
Y = Xb.
From Equation 12.2.6 we know that b
XtY, so we can rewrite
Y
X
HY,
(xtx)~‘xty
(12.2.11)
(12.2.12)
(12.2.13)
where H = X (xTx) ‘ XT is appropriately named the hat matrix because it “puts the hat on Y”.
The hat matrix is very important in later courses. Some facts about H are
• H is a symmetric square matrix, of dimension n x n.
• The diagonal entries h,, satisfy 0 < /;„■ < 1 (compare to Equation 11.5.2).
• The trace is tr(H) = p.
• H is idempotent (also known as a projection matrix ) which means that H2 = H. The same is
true of I - H.
292
CHAPTER 12. MULTIPLE LINEAR REGRESSION
Now let us write a column vector xo = (xio, X2o)T to denote given values of the explanatory variables
Girth = xio and Height = A'20 ■ These values may match those of the collected data, or they may
be completely new values not observed in the original data set. We may use the parameter estimates
to find T(xo), which will give us
1 . an estimate of fi(x 0), the mean value of a future observation at Xq, and
2. a prediction for T(xo), the actual value of a future observation at xo.
We can represent T(xo) by the matrix equation
T(x0) = xjb, (12.2.14)
which is just a fancy way to write
T(*io,X2o) = b0 + bixw + b2x2o- (12.2.15)
Example 12.4. If we wanted to predict the average volume of black cherry trees that have Girth
= 15 in and are Height = 77 ft tall then we would use the estimate
ju( 15, 77) = - 58 + 4.7(15) + 0.3(77),
^35.6 ft3.
We would use the same estimate Y = 35.6 to predict the measured Volume of another black cherry
tree - yet to be observed - that has Girth = 15 in and is Height = 77 ft tall.
12.2.4 How to do it with R
The fitted values are stored inside trees, lm and may be accessed with the fitted function. We
only show the first five fitted values.
> fitted(trees . lm) [1 : 5]
1 2 3 4 5
4.837660 4.553852 4.816981 15.874115 19.869008
The syntax for general prediction does not change much from simple linear regression. The
computations are done with the predict function as described below.
The only difference from SLR is in the way we tell R the values of the explanatory variables
for which we want predictions. In SLR we had only one independent variable but in MLR we have
many (for the trees data we have two). We will store values for the independent variables in the
data frame new, which has two columns (one for each independent variable) and three rows (we
shall make predictions at three different locations).
> new <- data, frame (Girth = c(9.1, 11.6, 12.5), Height = c(69,
+ 74, 87))
We can view the locations at which we will predict:
> new
12.2. ESTIMATION AND PREDICTION
293
Girth Height
1 9.1 69
2 11.6 74
3 12.5 87
We continue just like we would have done in SLR.
> predictCtrees.li n, newdata = new )
1 2 3
8.264937 21.731594 30.3792Q5
Example 12.5. Using the trees data,
1. Report a point estimate of the mean Volume of a tree of Girth 9.1 in and Height 69 ft.
The fitted value for x\ =9.1 and xo = 69 is 8.3, so a point estimate would be 8.3 cubic feet.
2. Report a point prediction for and a 95% prediction interval for the Volume of a hypothetical
tree of Girth 12.5 in and Height 87 ft.
The fitted value for x\ = 12.5 and X2 = 87 is 30.4, so a point prediction for the Volume is
30.4 cubic feet.
12.2.5 Mean Square Error and Standard Error
The residuals are given by
E = Y- Y = Y-HY = (I- H)Y. (12.2.16)
Now we can use Theorem 7.34 to see that the residuals are distributed
E ~ mvnormfmean = 0, sigma = <x2(I - H)), (12.2.17)
since (I - H)X|3 = X(3 - X|3 = 0 and (I - H) (cr2I) (I - H)T = cr2(I - H)2 = cr2(I - H). The sum
of squared errors SSE is just
SSE = EtE = Yt(I - H)(I - H)Y = YT(I - H)Y. (12.2.18)
Recall that in SLR we had two parameters (J3q and j3\) in our regression model and we estimated
cr2 with s1 - SSE/(n - 2). In MLR, we have p + 1 parameters in our regression model and we
might guess that to estimate cr2 we would use the mean square error S 2 defined by
52_ SSE
n-ip+iy
(12.2.19)
That would be a good guess. The residual standard error is S — Vs2.
294
CHAPTER 12. MULTIPLE LINEAR REGRESSION
12.2.6 How to do it with R
The residuals are also stored with trees . Ira and may be accessed with the residuals function.
We only show the first five residuals.
> residuals(trees . lm) [1 : 5]
1 2 3 4 5
5 . 46234Q3 5.7461484 5.3830187 0.5258848 -1.0690084
The summary function output (shown later) lists the Residual Standard Error which is
just S = vS2. It is stored in the sigma component of the summary object.
> treesumry <- summary (trees . lm)
> treesumry$sigma
[1] 3.881832
For the trees data we find s ~ 3.882.
12.2.7 Interval Estimates of the Parameters
We showed in Section 12.2.1 that b = (xTx) XTY, which is really just a big matrix - namely
(X ' X) XT - multiplied by Y. It stands to reason that the sampling distribution of b would be
intimately related to the distribution of Y, which we assumed to be
Y ~ mvnorm (mean = X|3, sigma = <x2l) . (12.2.20)
Now recall Theorem 7.34 that we said we were going to need eventually (the time is now). That
proposition guarantees that
b ~ mvnorm ^mean = (3, sigma = cr2(xTx) (12.2.21)
since
Eb = (xTX) Xt(X|3) = P, (12.2.22)
and
Var(b) = (XTX)~‘ XT(cr2I)X (xTx)~‘ = cr2 (xTx)~‘ , (12.2.23)
the first equality following because the matrix (xrXj is symmetric.
There is a lot that we can glean from Equation 12.2.21. First, it follows that the estimator b
is unbiased (see Section 9.1). Second, the variances of bo , b\ , .... b„ are exactly the diagonal
elements of cr2 (xTX) , which is completely known except for that pesky parameter cr2. Third,
we can estimate the standard error of h, (denoted S /,( ) with the mean square error S (defined in
the previous section) multiplied by the corresponding diagonal element of (X ' X ) . Finally, given
estimates of the standard errors we may construct confidence intervals for /3, with an interval that
looks like
bi ± tff/2(df = n — p - 1)5 br (12.2.24)
The degrees of freedom for the Student’s t distribution2 are the same as the denominator of S2.
2We are taking great leaps over the mathematical details. In particular, we have yet to show that s2 has a chi-square
12.2. ESTIMATION AND PREDICTION
295
12.2.8 How to do it with R
To get confidence intervals for the parameters we need only use confint:
> confint (trees, lm)
(Intercept)
Girth
Height
2.5 %
-75.68226247
4.16683899
0.07264863
97.5 %
-40.2930554
5.2494820
0.6058538
For example, using the calculations above we say that for the regression model Volume
~Girth + Height we are 95% confident that the parameter fi\ lies somewhere in the interval
[4.2, 5.2],
12.2.9 Confidence and Prediction Intervals
We saw in Section 12.2.3 how to make point estimates of the mean value of additional observations
and predict values of future observations, but how good are our estimates? We need confidence and
prediction intervals to gauge their accuracy, and lucky for us the formulas look similar to the ones
we saw in SLR.
In Equation 12.2.14 we wrote T(xo) = xlb, and in Equation 12.2.21 we saw that
b ~ mvnorm ^mean = (3, sigma = <x2 (xTx) * j , (12.2.25)
The following is therefore immediate from Theorem 7.34:
T(xq) ~ mvnorm ^mean = xj|3, sigma = cr2xj (xTx) ' x0j . (12.2.26)
It should be no surprise that confidence intervals for the mean value of a future observation at the
location xq = [xio x2o • ■ ■ xpo] are given by
T(x0) ± tff/2(df = n-p-l)S ^(XTxr'xo. (12.2.27)
Intuitively, x2 (XTXj xo measures the distance of Xo from the center of the data. The degrees of
freedom in the Student’s t critical value are n — (p + 1) because we need to estimate p+ 1 parameters.
Prediction intervals for a new observation at xo are given by
7(x0) ±tff/2(df = n-p-l)S ^1 +xJ(XTXr1x0. (12.2.28)
The prediction intervals are wider than the confidence intervals, just as in Section 1 1.2.5.
distribution and we have not even come close to showing that bj and Sbi are independent. But these are entirely outside the
scope of the present book and the reader may rest assured that the proofs await in later classes. See C.R. Rao for more.
296
CHAPTER 12. MULTIPLE LINEAR REGRESSION
12.2.10 How to do it with R
The syntax is identical to that used in SLR, with the proviso that we need to specify values of the
independent variables in the data frame new as we did in Section 11.2.5 (which we repeat here for
illustration).
> new <- data, frame (Girth = c(9.1, 11.6, 12.5), Height = c(69,
+ 74, 87))
Confidence intervals are given by
> predict(trees.lm, newdata = new, interval = "confidence")
fit lwr upr
1 8.264937 5.7724® 1®. 75747
2 21.731594 2®. 1111® 23.35208
3 3®. 379205 26.90964 33.84877
Prediction intervals are given by
> predict(trees.lm, newdata = new, interval = "prediction")
fit lwr upr
1 8.264937 -0. 06814444 16.59802
2 21.731594 13.61657775 29.84661
3 30.379205 21.70364103 39.05477
As before, the interval type is decided by the interval argument and the default confidence
level is 95% (which can be changed with the level argument).
Example 12.6. Using the trees data,
1. Report a 95% confidence interval for the mean Volume of a tree of Girth 9.1 in and Height
69 ft.
The 95% Cl is given by [5.8, 10.8], so with 95% confidence the mean Volume lies somewhere
between 5.8 cubic feet and 10.8 cubic feet.
2. Report a 95% prediction interval for the Volume of a hypothetical tree of Girth 12.5 in and
Height 87 ft.
The 95% prediction interval is given by [26.9, 33.8], so with 95% confidence we may assert
that the hypothetical Volume of a tree of Girth 12.5 in and Height 87 ft would lie some-
where between 26.9 cubic feet and 33.8 feet.
12.3 Model Utility and Inference
12.3.1 Multiple Coefficient of Determination
We saw in Section 12.2.5 that the error sum of squares S S E can be conveniently written in MLR
as
SSE - Yt(I - H)Y.
(12.3.1)
12.3. MODEL UTILITY AND INFERENCE
297
It turns out that there are equally convenient formulas for the total sum of squares SSTO and the
regression sum of squares SSR. They are :
and
SSTO =Yt
(12.3.2)
SSR =Yt
(12.3.3)
(The matrix J is defined in Appendix E.5.) Immediately from Equations 12.3.1, 12.3.2, and 12.3.3
we get the A nova Equality
SSTO = SSE + SSR. (12.3.4)
(See Exercise 12.1.) We define the multiple coefficient of determination by the formula
, SSE
SSTO'
(12.3.5)
We interpret R2 as the proportion of total variation that is explained by the multiple regression
model. In MLR we must be careful, however, because the value of R2 can be artificially inflated by
the addition of explanatory variables to the model, regardless of whether or not the added variables
are useful with respect to prediction of the response variable. In fact, it can be proved that the
addition of a single explanatory variable to a regression model will increase the value of R2, no
matter how worthless the explanatory variable is. We could model the height of the ocean tides,
then add a variable for the length of cheetah tongues on the Serengeti plain, and our R2 would
inevitably increase.
This is a problem, because as the philosopher, Occam, once said: “causes should not be multi-
plied beyond necessity”. We address the problem by penalizing R2 when parameters are added to
? —2
the model. The result is an adjusted R which we denote by R .
—2
R
n — 1
~ P~
(12.3.6)
T —2
It is good practice for the statistician to weigh both R~ and R during assessment of model utility.
In many cases their values will be very close to each other. If their values differ substantially, or
if one changes dramatically when an explanatory variable is added, then (s)he should take a closer
look at the explanatory variables in the model.
12.3.2 How to do it with R
, —2
For the trees data, we can get R - and R from the summary output or access the values directly
by name as shown (recall that we stored the summary object in treesumry).
> treesumry$r. squared
[1] 0.94795
> treesumry $adj .r. squared
[1] 0.9442322
2 —2
High values of R~ and R such as these indicate that the model fits very well, which agrees with
what we saw in Figure 12.1.2.
298
CHAPTER 12. MULTIPLE LINEAR REGRESSION
12.3.3 Overall F- Test
Another way to assess the model’s utility is to to test the hypothesis
Hq : Pi — P2 = • • • = Pp — 0 versus H\ : at least one Pi 4 0.
The idea is that if all /?,’s were zero, then the explanatory variables X\, ... ,Xp would be worthless
predictors for the response variable Y. We can test the above hypothesis with the overall F statistic,
which in MLR is defined by
SSR/p
SSE/(n- p- 1)'
(12.3.7)
When the regression assumptions hold and under Ho, it can be shown that F ~ f(dfl = p. df2 =
n - p - 1). We reject Ho when F is large, that is, when the explained variation is large relative to
the unexplained variation.
12.3.4 How to do it with R
The overall F statistic and its associated p- value is listed at the bottom of the summary output,
or we can access it directly by name; it is stored in the fstatistic component of the summary
object.
> treesumry$ fstatistic
value numdf dendf
254.9723 2.0000 28.000®
For the trees data, we see that F = 254.972337410669 with a p- value < 2 . 2e-16. Conse-
quently we reject Hq, that is, the data provide strong evidence that not all s are zero.
12.3.5 Student’s t Tests
We know that
b ~ mvnorm ^mean = (3, sigma = cr2 (xTx) * j (12.3.8)
and we have seen how to test the hypothesis Hq : P\ = Pi = • • • = Pp = 0, but let us now consider
the test
H0 : Pi = 0 versus Hx : Pi 4 0, (12.3.9)
where pt is the coefficient for the ;th independent variable. We test the hypothesis by calculating a
statistic, examining it’s null distribution, and rejecting Hq if the p-value is small. If Ho is rejected,
then we conclude that there is a significant relationship between Y and x, in the regression model
Y ~ (X[ xp). This last part of the sentence is very important because the significance of the
variable x, sometimes depends on the presence of other independent variables in the model3.
3 In other words, a variable might be highly significant one moment but then fail to be significant when another variable is
added to the model. When this happens it often indicates a problem with the explanatory variables, such as multicollinearity.
See Section 12.9.3.
12.4. POLYNOMIAL REGRESSION
299
To test the hypothesis we go to find the sampling distribution of b the estimator of the corre-
sponding parameter when the null hypothesis is true. We saw in Section 12.2.7 that
Tt =
bj-Pi
Sbi
(12.3.10)
has a Student’s t distribution with n - (p + 1) degrees of freedom. (Remember, we are estimating
p + 1 parameters.) Consequently, under the null hypothesis //0 : fi, - 0 the statistic t, = bJS bj has
a t(df = n - p - 1) distribution.
12.3.6 How to do it with R
The Student’s t tests for significance of the individual explanatory variables are shown in the
summary output.
> treesumry
Call:
lm(formula = Volume ~ Girth + Height, data = trees)
Residuals:
Min IQ Median 3Q Max
-6.4065 -2.6493 -0.2876 2.2003 8.4847
Coefficients :
Estimate
(Intercept) -57.9877
Girth 4.7082
Height 0.3393
Signif. codes: 0 '***' 0.001 '**' 0.01 0.05 0.1 ' ' 1
Residual standard error: 3.882 on 28 degrees of freedom
Multiple R-squared: 0.948, Adjusted R-squared: 0.9442
F-statistic: 255 on 2 and 28 DF, p-value: < 2.2e-16
We see from the /7-values that there is a significant linear relationship between Volume and
Girth and between Volume and Height in the regression model Volume ~Girth + Height.
Further, it appears that the Intercept is significant in the aforementioned model.
Std. Error t value Pr(>|t|)
8.6382 -6.713 2.75e-07
0.2643 17.816 < 2e-16
0.1302 2.607 0.0145
12.4 Polynomial Regression
12.4.1 Quadratic Regression Model
In each of the previous sections we assumed that p was a linear function of the explanatory vari-
ables. For example, in SLR we assumed that p(x) = fio + (i \ x, and in our previous MLR examples
we assumed p(x i,X2) = /?o+/3iXi - In every case the scatterplots indicated that our assumption
Volume
300
CHAPTER 12. MULTIPLE LINEAR REGRESSION
8 10 12 14 16 18 20
Girth
Figure 12.4.1: Scatterplot of Volume versus Girth for the trees data
12.4. POLYNOMIAL REGRESSION
301
was reasonable. Sometimes, however, plots of the data suggest that the linear model is incomplete
and should be modified.
For example, let us examine a scatterplot of Volume versus Girth a little more closely. See
Figure 12.4.1. There might be a slight curvature to the data; the volume curves ever so slightly
upward as the girth increases. After looking at the plot we might try to capture the curvature with
a mean response such as
ju(xi) = fio +P\X\ +Pix\. (12.4.1)
The model associated with this choice of p is
Y — Po + P\X\ + Pix\ + £■ (12.4.2)
The regression assumptions are the same. Almost everything indeed is the same. In fact, it is still
called a “linear regression model”, since the mean response ) u is linear in the parameters Pq, Pi,
and Pi.
However, there is one important difference. When we introduce the squared variable in the
model we inadvertently also introduce strong dependence between the terms which can cause sig-
nificant numerical problems when it comes time to calculate the parameter estimates. Therefore,
we should usually rescale the independent variable to have mean zero (and even variance one if we
wish) before fitting the model. That is, we replace the x,’s with x, - x (or (x,- - x)/s) before fitting
the model4.
How to do it with R
There are multiple ways to fit a quadratic model to the variables Volume and Girth using R.
1. One way would be to square the values for Girth and save them in a vector Girthsq. Next,
fit the lineal' model Volume ~Girth + Girthsq.
2. A second way would be to use the insulate function in R, denoted by I:
Volume ~ Girth + I(GirthA2)
The second method is shorter than the first but the end result is the same. And once we
calculate and store the fitted model (in, say, treesquad. lm) all of the previous comments
regarding R apply.
3. A third and “right” way to do it is with orthogonal polynomials:
Volume ~ poly(Girth, degree = 2)
See ?poly and ?cars for more information. Note that we can recover the approach in 2 with
poly (Girth, degree = 2, raw = TRUE).
4Rescaling the data gets the job done but a better way to avoid the multicollinearity introduced by the higher order terms
is with orthogonal polynomials, whose coefficients are chosen just right so that the polynomials are not correlated with each
other. This is beginning to linger outside the scope of this book, however, so we will content ourselves with a brief mention
and then stick with the rescaling approach in the discussion that follows. A nice example of orthogonal polynomials in
action can be run with example (cars).
302
CHAPTER 12. MULTIPLE LINEAR REGRESSION
Example 12.7. We will fit the quadratic model to the trees data and display the results with
summary, being careful to rescale the data before fitting the model. We may rescale the Girth
variable to have zero mean and unit variance on-the-fly with the scale function.
> treesquad. lm <- lm(Volume ~ scale(Girth) + I (scale (Girth) A 2) ,
+ data = trees)
> summary (treesquad. lm)
Call:
lmfformula = Volume ~ scale(Girth) + I (scale (Girth) A2) , data = trees)
Residuals:
Min IQ Median 3Q Max
-5.4889 -2.4293 -0.3718 2.0764 7.6447
Coefficients :
Estimate Std. Error t value Pr(>|t|)
(Intercept) 27.7452 0.8161 33.996 < 2e-16 ***
scale (Girth) 14.5995 0.6773 21.557 < 2e-16 ***
I (scale (Girth) A 2) 2.5067 0.5729 4.376 0.000152 ***
Signif. codes: 0 '***' 0.001 '**' 0.01 0.05 0.1 ' ' 1
Residual standard error: 3.335 on 28 degrees of freedom
Multiple R-squared: 0.9616, Adjusted R-squared: 0.9588
F-statistic: 350.5 on 2 and 28 DF, p-value: < 2.2e-16
We see that the F statistic indicates the overall model including Girth and GirthA2 is signif-
icant. Further, there is strong evidence that both Girth and GirthA2 are significantly related to
Volume. We may examine a scatterplot together with the fitted quadratic function using the lines
function, which adds a line to the plot tracing the estimated mean response.
> plot (Volume ~ scale(Girth) , data = trees)
> lines(fitted(treesquad. lm) ~ scale(Girth) , data = trees)
The plot is shown in Figure 12.4.2. Pay attention to the scale on the x-axis: it is on the scale of
the transformed Girth data and not on the original scale.
Remark 12.8. When a model includes a quadratic term for an independent variable, it is customary
to also include the linear term in the model. The principle is called parsimony. More generally, if
the researcher decides to include x"’ as a term in the model, then (s)he should also include all lower
order terms x, x2, . . . ,xf" 1 in the model.
We do estimation/prediction the same way that we did in Section 12.2.3, except we do not need
a Height column in the dataframe new since the variable is not included in the quadratic model.
> new <- data, frame (Girth = c(9.1, 11.6, 12.5))
> predict(treesquad.lm, newdata = new, interval = "prediction")
Volume
12.4. POLYNOMIAL REGRESSION
303
scale(Girth)
Figure 12.4.2: A quadratic model for the trees data
304
CHAPTER 12. MULTIPLE LINEAR REGRESSION
fit
1 11.56982
2 2®. 30615
3 25.9229®
lwr upr
4.347426 18.79221
13.29905® 27.31325
18.972934 32.87286
The predictions and intervals are slightly different from what they were previously. Notice that
it was not necessary to rescale the Girth prediction data before input to the predict function; the
model did the rescaling for us automatically.
Remark 12.9. We have mentioned on several occasions that it is important to rescale the explana-
tory variables for polynomial regression. Watch what happens if we ignore this advice:
> summary(lm(Volume ~ Girth + I(GirthA2) , data = trees))
Call:
lm(formula = Volume ~ Girth + I(GirthA2), data = trees)
Residuals:
Min IQ Median 3Q Max
-5.4889 -2.4293 -Q.3718 2. 0764 7.6447
Coefficients :
Estimate
(Intercept) 16.78627
Girth -2.09214
I(GirthA2) 0.25454
Signif. codes: 0 '*
Std. Error t value
11.22282 0.961
1.64734 -1.270
0.05817 4.376
*' 0.001 '**' 0.01 '
Pr C> I t | )
0.344728
0.214534
0.000152 ***
' 0.05 ' . ' 0.1 ' ' 1
Residual standard error: 3.335 on 28 degrees of freedom
Multiple R-squared: 0.9616, Adjusted R-squared: 0.9588
F-statistic: 350.5 on 2 and 28 DF, p-value: < 2.2e-16
Now nothing is significant in the model except GirthA2. We could delete the Intercept and
Girth from the model, but the model would no longer be parsimonious. A novice may see the
output and be confused about how to proceed, while the seasoned statistician recognizes immedi-
ately that Girth and GirthA2 are highly correlated (see Section 12.9.3). The only remedy to this
ailment is to rescale Girth, which we should have done in the first place.
In Example 12.14 of Section 12.7 we investigate this issue further.
12.5 Interaction
In our model for tree volume there have been two independent variables: Girth and Height. We
may suspect that the independent variables are related, that is, values of one variable may tend to
influence values of the other. It may be desirable to include an additional term in our model to try
and capture the dependence between the variables. Interaction terms are formed by multiplying
one (or more) explanatory variable(s) by another.
12.5. INTERACTION
305
Example 12.10. Perhaps the Girth and Height of the tree interact to influence the its Volume;
we would like to investigate whether the model (Girth = x\ and Height = xT)
Y — fia + (3\Xi + P2X2 + e (12.5.1)
would be significantly improved by the model
Y - j Bq + p\x\ + P2X2 +P\aX\X2 + e, (12.5.2)
where the subscript 1 : 2 denotes that P\-2 is a coefficient of an interaction term between xi and .ty.
What does it mean? Consider the mean response g(x\ , X2) as a function of xy.
T(x2) - (Po + P\x\) + P2X2- (12.5.3)
This is a linear function of X2 with slope /T. As x\ changes, the y-intercept of the mean response in
X2 changes, but the slope remains the same. Therefore, the mean response in xo is represented by a
collection of parallel lines all with common slope /?2-
Now think about what happens when the interaction term P\ax\x2 is included. The mean re-
sponse in X2 now looks like
M*2) = ( Po +P\xi) + (P2 +P\-.2X\)X2- (12.5.4)
In this case we see that not only the y-intercept changes when jcj varies, but the slope also changes in
X\ . Thus, the interaction term allows the slope of the mean response in X2 to increase and decrease
as xi varies.
How to do it with R
There are several ways to introduce an interaction term into the model.
1. Make a new variable prod <-Girth '"Height, then include prod in the model formula
Volume ~Girth + Height + prod. This method is perhaps the most transparent, but it
also reserves memory space unnecessarily.
2. Once can construct an interaction term directly in R with a colon “ : ”. For this example, the
model formula would look like Volume ~Girth + Height + Girth: Height.
For the trees data, we fit the model with the interaction using method two and see if it is signifi-
cant:
> treesint.lm <- lm(Volume ~ Girth + Height + Girth : Height ,
+ data = trees)
> summaryCtreesint . lm)
Call:
lm(formula = Volume ~ Girth + Height + Girth: Height, data = trees)
Residuals:
306
CHAPTER 12. MULTIPLE LINEAR REGRESSION
Min IQ Median 3Q Max
-6.5821 -1.0673 0.3026 1.5641 4.6649
Coefficients :
Estimate Std. Error t value
(Intercept) 69.39632
Girth -5.85585
Height -1.29708
Girth:Height 0.13465
Signif. codes: 0 '**
23.83575 2.911
1.92134 -3.048
0.30984 -4.186
0.02438 5.524
' 0.001 '**' 0.01 '
Pr (> I t | )
0.00713 **
0.00511 **
0.00027 ***
7 . 48e-06 ***
*' 0.05 ' . ' 0.1 ' ' 1
Residual standard error: 2.709 on 27 degrees of freedom
Multiple R-squared: 0.9756, Adjusted R-squared: 0.9728
F-statistic: 359.3 on 3 and 27 DF, p-value: < 2.2e-16
We can see from the output that the interaction term is highly significant. Further, the estimate
b i;2 is positive. This means that the slope of p(x2) is steeper for bigger values of Girth. Keep in
mind: the same interpretation holds for g(x i); that is, the slope of p(x\ ) is steeper for bigger values
of Height.
For the sake of completeness we calculate confidence intervals for the parameters and do pre-
diction as before.
> confint(treesint.lm)
2.5 % 97.5 %
(Intercept) 20.48938699 118.3032441
Girth -9.79810354 -1.9135923
Height -1.93282845 -0.6613383
Girth: Height 0.08463628 0.1846725
> new <- data, frame (Girth = c(9.1, 11.6, 12.5), Height = c(69,
+ 74, 87))
> predict (treesint . lm, newdata = new, interval = "prediction")
fit lwr upr
1 11.15884 5.236341 17.08134
2 21.07164 15.394628 26.74866
3 29.78862 23.721155 35.85608
Remark 12.11. There are two other ways to include interaction terms in model formulas. For
example, we could have written Girth ’-Height or even (Girth + Height) A 2 and both would
be the same as Girth + Height + Girth: Height.
These examples can be generalized to more than two independent variables, say three, four, or even
more. We may be interested in seeing whether any pairwise interactions are significant. We do this
with a model formula that looks something like y ~ (xl + x2 + x3 + x4)A2.
12.6. QUALITATIVE EXPLANATORY VARIABLES
307
12.6 Qualitative Explanatory Variables
We have so far been concerned with numerical independent variables taking values in a subset of
real numbers. In this section, we extend our treatment to include the case in which one of the
explanatory variables is qualitative, that is, a factor. Qualitative variables take values in a set of
levels, which may or may not be ordered. See Section 3.1.2.
Note. The trees data do not have any qualitative explanatory variables, so we will construct one
for illustrative purposes3. We will leave the Girth variable alone, but we will replace the variable
Height by a new variable Tall which indicates whether or not the cherry tree is taller than a
certain threshold (which for the sake of argument will be the sample median height of 76 ft). That
is. Tall will be defined by
Tall =
yes,
no,
if Height > 76,
if Height < 76.
(12.6.1)
We can construct Tall very quickly in R with the cut function:
> trees$Tall <- cut(trees$Height , breaks = c(-Inf, 76, Inf),
+ labels = c("no", "yes"))
> trees$Tall[l: 5]
[1] no no no no yes
Levels: no yes
Note that Tall is automatically generated to be a factor with the labels in the correct order. See
?cut for more.
Once we have Tall, we include it in the regression model just like we would any other variable.
It is handled internally in a special way. Define a “dummy variable” Tallyes that takes values
Tallyes
1 1 Ta 1 1 = yes,
otherwise.
(12.6.2)
That is, Tallyes is an indicator variable which indicates when a respective tree is tall. The model
may now be written as
Volume = /?o + /liGirth + /TTallyes + e. (12.6.3)
Let us take a look at what this definition does to the mean response. Trees with Tall = yes will
have the mean response
/r(Girth) = (/?o + Pi) + ySiGirth, (12.6.4)
while trees with Tall = no will have the mean response
/r(Girth) = /?o + P\ Girth. (12.6.5)
In essence, we are fitting two regression lines: one for tall trees, and one for short trees. The
regression lines have the same slope but they have different y intercepts (which are exactly |/T| far
apart).
5This procedure of replacing a continuous variable by a discrete/qualitative one is called binning , and is almost never
the right thing to do. We are in a bind at this point, however, because we have invested this chapter in the trees data and
I do not want to switch mid-discussion. I am currently searching for a data set with pre-existing qualitative variables that
also conveys the same points present in the trees data, and when I find it I will update this chapter accordingly.
308
CHAPTER 12. MULTIPLE LINEAR REGRESSION
How to do it with R
The important thing is to double check that the qualitative variable in question is stored as a factor.
The way to check is with the class command. For example,
> class(trees$Tall)
[1] "factor"
If the qualitative variable is not yet stored as a factor then we may convert it to one with the
factor command. See Section 3.1.2. Other than this we perform MLR as we normally would.
> treesduimy.lm <- lm(Volume ~ Girth + Tall, data = trees)
> summaryCtreesdummy.il n)
Call:
lmfformula = Volume ~ Girth + Tall, data = trees)
Residuals:
Min IQ Median
-5.7788 -3.1718 8.4888
3Q
2.6737
Max
18.8619
Coefficients :
Estimate Std
. Error
t value
Pr(> | t | )
(Intercept) -34.1652
3.2438
-18.53
3.82e-ll
Girth 4.6988
8.2652
17.72
< 2e-16
Tallyes 4.3872
1.6388
2.63
8.8137
Signif. codes: 8 '***'
8.881 ”
8.81
8.85
Residual standard error: 3.875 on 28 degrees of freedom
Multiple R-squared: 8.9481, Adjusted R-squared: 8.9444
F-statistic: 255.9 on 2 and 28 DF, p-value: < 2.2e-16
From the output we see that all parameter estimates are statistically significant and we conclude
that the mean response differs for trees with Tall = yes and trees with Tall = no.
Remark 12.12. We were somewhat disingenuous when we defined the dummy variable Tallyes
because, in truth, R defines Tallyes automatically without input from the user6. Indeed, the author
fit the model beforehand and wrote the discussion afterward with the knowledge of what R would
do so that the output the reader saw would match what (s)he had previously read. The way that R
handles factors internally is part of a much larger topic concerning contrasts, which falls outside
the scope of this book. The interested reader should see Neter et al [67] or Fox [28] for more.
Remark 12.13. In general, if an explanatory variable foo is qualitative with n levels barl, bar2,
. . . , barn then R will by default automatically define n — 1 indicator variables in the following way:
fl, if foo = ’’bar 2”, fl, if foo = "barn”,
foobar2 = < , . . . , foobarn = <
10, otherwise. 10, otherwise.
6That is. R by default handles contrasts according to its internal settings which may be customized by the user for fine
control. Given that we will not investigate contrasts further in this book it does not serve the discussion to delve into those
settings, either. The interested reader should check ?contrasts for details.
12.6. QUALITATIVE EXPLANATORY VARIABLES
309
Girth
Figure 12.6.1: A dummy variable model for the trees data
The level barl is represented by foobar2 = • • • = foobarn = 0. We just need to make sure that
foo is stored as a factor and R will take care of the rest.
Graphing the Regression Lines
We can see a plot of the two regression lines with the following mouthful of code.
> treesTall <- split(trees, trees$Tall)
> treesTall[["yes"]]$Fit <- predict(treesduimy.lm, treesTall [ ["yes"] ] )
> treesTall[["no"]]$Fit <- predict (treesdummy . lm, treesTall [["no"]])
> plot (Volume ~ Girth, data = trees, type = "n")
> points(Volume ~ Girth, data = treesTall [ ["yes"] ] , pch = 1)
> points(Volume ~ Girth, data = treesTall [ ["no"] ] , pch = 2)
> lines(Fit ~ Girth, data = treesTall [ ["yes"] ] )
> lines(Fit ~ Girth, data = treesTall[["no"]])
It may look intimidating but there is reason to the madness. First we split the trees data into
two pieces, with groups determined by the Tall variable. Next we add the Fitted values to each
piece via predict. Then we set up a plot for the variables Volume versus Girth, but we do not
310
CHAPTER 12. MULTIPLE LINEAR REGRESSION
plot anything yet (type = n) because we want to use different symbols for the two groups. Next
we add points to the plot for the Tall = yes trees and use an open circle for a plot character
(pch = 1), followed by points for the Tall = no trees with a triangle character (pch = 2).
Finally, we add regression lines to the plot, one for each group.
There are other - shorter - ways to plot regression lines by groups, namely the scatterplot
function in the car [30] package and the xyplot function in the lattice package. We elected to
introduce the reader to the above approach since many advanced plots in R are done in a similar,
consecutive fashion.
12.7 Partial F Statistic
We saw in Section 12.3.3 how to test Hq : /3q = = ■ ■ ■ = f3p = 0 with the overall F statistic and
we saw in Section 12.3.5 how to test Hq : [3, = 0 that a particular coefficient /?, is zero. Sometimes,
however, we would like to test whether a certain part of the model is significant. Consider the
regression model
Y - A) + P\X\ + • ■ • + PjXj + Pj+iXj+i + • ■ ■ + PpXp + e, (12.7.1)
where j > 1 and p > 2. Now we wish to test the hypothesis
Hq : Pj+ 1 - fij+2 - • • • - PP - 0
(12.7.2)
versus the alternative
H\ : at least one of (ij+ \ , fij+i, , ■ ■ ■ ,fiP + 0. (12.7.3)
The interpretation of Hq is that none of the variables Xj+ 1 .... ,xp is significantly related to Y and
the interpretation of H\ is that at least one of Xj+ 1, . . . ,xp is significantly related to Y . In essence,
for this hypothesis test there are two competing models under consideration:
the full model: y = /3q + fi\X\ H + fipxp + e, (12.7.4)
the reduced model: y = /3q + fi\X\ + ■ ■ ■ + \SjXj + e, (12.7.5)
Of course, the full model will always explain the data better than the reduced model, but does the
full model explain the data significantly better than the reduced model? This question is exactly
what the partial F statistic is designed to answer.
We first calculate SSEf , the unexplained variation in the full model, and S S Er. the unexplained
variation in the reduced model. We base our test on the difference SSEr-SSEf which measures
the reduction in unexplained variation attributable to the variables Xj+ 1, . . . ,xp. In the full model
there are p + 1 parameters and in the reduced model there are j + 1 parameters, which gives a
difference of p - j parameters (hence degrees of freedom). The partial F statistic is
(SSEr-SSEf)/(p-j)
SSEfl(n-p- 1)
(12.7.6)
It can be shown when the regression assumptions hold under Hq that the partial F statistic has an
f(dfl = p — j, df2 — n — p — 1) distribution. We calculate the /;- value of the observed partial F
statistic and reject Hq if the p- value is small.
12.7. PARTIAL F STATISTIC
311
How to do it with R
The key ingredient above is that the two competing models are nested in the sense that the reduced
model is entirely contained within the complete model. The way to test whether the improvement
is significant is to compute lm objects both for the complete model and the reduced model then
compare the answers with the anova function.
Example 12.14. For the trees data, let us fit a polynomial regression model and for the sake of
argument we will ignore our own good advice and fail to rescale the explanatory variables.
> trees full . lm <- lm(Volume ~ Girth + I(GirthA2) + Height +
+ I(HeightA2) , data = trees )
> summary (treesfull .lm)
Call:
lm(formula = Volume ~ Girth + I(GirthA2) + Height + I(HeightA2),
data = trees)
Residuals:
Min IQ Median 3Q Max
-4.368 -1.67® -0.158 1.792 4.358
Coefficients :
Estimate
Std. Error
t value
Pr(> | t | )
(Intercept)
-0.955101
63.013630
-0.015
0.988
Girth
-2.796569
1.468677
-1.904
0.068 .
I(GirthA2)
0.265446
0.051689
5.135
2 . 35e-05 *
Height
0.119372
1.784588
0.067
0.947
I(HeightA2)
0.001717
0.011905
0.144
0.886
Signif. codes: 0
■" 0.001 '*■
0.01 ’
0.05 ' .
' 0.1
Residual standard error: 2.674 on 26 degrees of freedom
Multiple R-squared: 0.9771, Adjusted R-squared: 0.9735
F-statistic: 277 on 4 and 26 DF, p-value: < 2.2e-16
In this ill-formed model nothing is significant except Girth and GirthA2. Let us continue
down this path and suppose that we would like to try a reduced model which contains nothing but
Girth and GirthA2 (not even an Intercept). Our two models are now
the full model: Y = fio + f3\X\ + /?2*i + P1X2 + @4X2 + e,
the reduced model: Y = fl\X\ + /L-tf + e,
We fit the reduced model with lm and store the results:
> treesreduced. lm <- lm(Volume ~ -1 + Girth + I(GirthA2) , data = trees)
To delete the intercept from the model we used - 1 in the model formula. Next we compare the
two models with the anova function. The convention is to list the models from smallest to largest.
312
CHAPTER 12. MULTIPLE LINEAR REGRESSION
> anova(treesreduced. lm, trees full . lm)
Analysis of Variance Table
Model 1: Volume ~ -1 + Girth + I(GirthA2)
Model 2: Volume ~ Girth + I(GirthA2) + Height + I(HeightA2)
Res.Df RSS Df Sum of Sq F Pr(>F)
1 29 321.65
2 26 185.86 3 135.79 6.3319 0.002279 **
Signif. codes: 0 '***' 0.001 '**' 0.01 0.05 0.1 ' ' 1
We see from the output that the complete model is highly significant compared to the model
that does not incorporate Height or the Intercept. We wonder (with our tongue in our cheek) if
the HeightA2 term in the full model is causing all of the trouble. We will fit an alternative reduced
model that only deletes HeightA2.
> treesreduced2 .lm <- lm(Volume ~ Girth + I(GirthA2 ) + Height,
+ data = trees)
> anova(treesreduced2 . lm, trees full . lm)
Analysis of Variance Table
Model 1: Volume ~ Girth + I(GirthA2) + Height
Model 2: Volume ~ Girth + I(GirthA2) + Height + I(HeightA2)
Res.Df RSS Df Sum of Sq F Pr(>F)
1 27 186.01
2 26 185.86 1 0.14865 0.0208 0.8865
In this case, the improvement to the reduced model that is attributable to HeightA2 is not
significant, so we can delete HeightA2 from the model with a clear conscience. We notice that the
p- value for this latest partial F test is 0.8865, which seems to be remarkably close to the p- value we
saw for the univariate t test of Height A 2 at the beginning of this example. In fact, the p-values are
exactly the same. Perhaps now we gain some insight into the true meaning of the univariate tests.
12.8 Residual Analysis and Diagnostic Tools
We encountered many, many diagnostic measures for simple linear regression in Sections 1 1.4 and
11.5. All of these are valid in multiple linear regression, too, but there are some slight changes
that we need to make for the multivariate case. We list these below, and apply them to the trees
example.
Shapiro-Wilk, Breusch-Pagan, Durbin-Watson: unchanged from SLR, but we are now equipped
to talk about the Shapiro-Wilk test statistic for the residuals. It is defined by the formula
aTE*
ETE’
W =
(12.8.1)
12.9. ADDITIONAL TOPICS
313
where E* is the sorted residuals and aixn is defined by
mTV 1
(12.8.2)
VmTV 1 V 'm
where mnxi and Vnxn are the mean and covariance matrix, respectively, of the order statistics
from an mvnorm (mean = 0, sigma = I) distribution.
Leverages: are defined to be the diagonal entries of the hat matrix H (which is why we called
them ha in Section 12.2.3). The sum of the leverages is tr(H) = p + 1. One rule of thumb
considers a leverage extreme if it is larger than double the mean leverage value, which is
2 (p + 1 )/n, and another rule of thumb considers leverages bigger than 0.5 to indicate high
leverage, while values between 0.3 and 0.5 indicate moderate leverage.
Standardized residuals: unchanged. Considered extreme if \Ri\ > 2.
Studentized residuals: compared to a t( df = n — p — 2) distribution.
DF BETAS : The formula is generalized to
where c jj is the /h diagonal entry of (XTX) 1 . Values larger than one for small data sets or
2/ yfn for large data sets should be investigated.
DFFITS: unchanged. Larger than one in absolute value is considered extreme.
Cook’s D: compared to an f(df 1 = p + 1, df2 = n — p — 1) distribution. Observations falling
higher than the 50th percentile are extreme.
Note that plugging the value p = 1 into the formulas will recover all of the ones we saw in Chapter
11.
12.9 Additional Topics
12.9.1 Nonlinear Regression
We spent the entire chapter talking about the trees data, and all of our models looked like Volume
~Girth + Height or a variant of this model. But let us think again: we know from elementary
school that the volume of a rectangle is V = Iwh and the volume of a cylinder (which is closer to
what a black cherry tree looks like) is
where r and d represent the radius and diameter of the tree, respectively. With this in mind, it would
seem that a more appropriate model for p might be
(DF BETAS )m =
j = 0,...p, i =
(12.8.3)
V = nr2h or V — 4ndh,
(12.9.1)
p(x\,x2) =
(12.9.2)
314
CHAPTER 12. MULTIPLE LINEAR REGRESSION
where [3\ and [T are parameters to adjust for the fact that a black cherry tree is not a perfect cylinder.
How can we fit this model? The model is not linear in the parameters any more, so our linear
regression methods will not work. . . or will they? In the trees example we may take the logarithm
of both sides of Equation 12.9.2 to get
jj*(X],x2) = In [pi(x\ , x2)] = InySo + /3i lnxi +/T lnx2, (12.9.3)
and this new model p* is linear in the parameters /)* = ln/?o, = P\ and f 3 * - /T. We can use what
we have learned to fit a linear model log(Volume)~log(Girth)+ log(Height), and everything
will proceed as before, with one exception: we will need to be mindful when it comes time to make
predictions because the model will have been fit on the log scale, and we will need to transform our
predictions back to the original scale (by exponentiating with exp) to make sense.
> treesNonlin. lm <- lm(log (Volume) ~ log(Girth) + log (Height) ,
+ data = trees)
> summary (treesNonlin. lm)
Call:
lm(formula = log(Volume) ~ log(Girth) + log(Height), data = trees)
Residuals:
Min IQ
Median
3Q
Max
-0.168561 -0.048488
0.002431 0
.063637
0.129223
Coefficients :
Estimate
Std. Error
t value
Pr(> | t | )
(Intercept) -6.63162
0.79979
-8.292
5.06e-09
log(Girth) 1.98265
0.07501
26.432
< 2e-16
log(Height) 1.11712
0.20444
5.464
7 . 81e-06
Signif. codes: 0 '
■*' 0.001 '*
*' 0.01
0.05
Residual standard error: 0.08139 on 28 degrees of freedom
Multiple R-squared: 0.9777, Adjusted R-squared: 0.9761
F-statistic: 613.2 on 2 and 28 DF, p-value: < 2.2e-16
i —2
This is our best model yet (judging by IP and R ), all of the parameters are significant, it is
simpler than the quadratic or interaction models, and it even makes theoretical sense. It rarely gets
any better than that.
We may get confidence intervals for the parameters, but remember that it is usually better to
transform back to the original scale for interpretation purposes :
> exp (con fint (treesNonlin. lm))
2.5 % 97.5 %
(Intercept) 0.0002561078 0.006783093
log (Girth) 6.2276411645 8.468066317
log (Height) 2.0104387829 4.645475188
12.9. ADDITIONAL TOPICS
315
(Note that we did not update the row labels of the matrix to show that we exponentiated and
so they are misleading as written.) We do predictions just as before. Remember to transform the
response variable back to the original scale after prediction.
> new <- data, frame (Girth = c(9.1, 11.6, 12.5), Height = c(69,
+ 74, 87))
> exp (predict (treesNonlin. lm, newdata = new, interval = "confidence"))
fit lwr upr
1 11.90117 11.25908 12.57989
2 20.82261 20.14652 21.52139
3 28.93317 27.03755 30.96169
The predictions and intervals are slightly different from those calculated earlier, but they are
close. Note that we did not need to transform the Girth and Height arguments in the dataframe
new. All transformations are done for us automatically.
12.9.2 Real Nonlinear Regression
We saw with the trees data that a nonlinear model might be more appropriate for the data based
on theoretical considerations, and we were lucky because the functional form of p allowed us to
take logarithms to transform the nonlinear model to a linear one. The same trick will not work in
other circumstances, however. We need techniques to fit general models of the form
Y = p(X) + e, (12.9.4)
where p is some crazy function that does not lend itself to linear transformations.
There are a host of methods to address problems like these which are studied in advanced
regression classes. The interested reader should see Neter et al [67] or Tabachnick and Fidell [83].
It turns out that John Fox has posted an Appendix to his book [29] which discusses some of the
methods and issues associated with nonlinear regression; see
http : //cran . r-pro j ect . org/doc/contrib/Fox- Companion/ appendix . html
Here is an example of how it works, based on a question from R-help.
> set . seed(l)
> x <- seq(from = Q, to - 1666, length. out = 266)
> y <- 1 + 2 * (sin((2 * pi * x/366) - 3))A2 + rnorm(266 , sd = 2)
> plot (x , y)
> acc.nls <- nls(y ~ a + b * (sin((2 * pi * x/366) - c))A2,
+ start = list(a = 6.9, b = 2.3, c = 2.9))
> summary (acc .nls)
Formula: y ~ a + b * (sin((2 * pi * x/360) - c))A2
Parameters :
Estimate Std. Error t value Pr(>|t|)
a 0.95884 0.23097 4.151 4.92e-05 ***
316
CHAPTER 12. MULTIPLE LINEAR REGRESSION
b 2.22868 (3.37114 6.0(35 9.07e-09 ***
c 3.04343 0.08434 36.084 < 2e-16 ***
Signif. codes: 0 '***' 0.001 '**' 0.01 0.05 0.1 ' ' 1
Residual standard error: 1.865 on 197 degrees of freedom
Number of iterations to convergence: 3
Achieved convergence tolerance: 6.554e-08
12.9.3 Multicollinearity
A multiple regression model exhibits multicollinearity when two or more of the explanatory vari-
ables are substantially correlated with each other. We can measure multicollinearity by having one
of the explanatory play the role of “dependent variable” and regress it on the remaining explanatory
variables. The the R 2 of the resulting model is near one, then we say that the model is multicollinear
or shows multicollinearity.
Multicollinearity is a problem because it causes instability in the regression model. The insta-
bility is a consequence of redundancy in the explanatory variables: a high R2 indicates a strong
dependence between the selected independent variable and the others. The redundant information
inflates the variance of the parameter estimates which can cause them to be statistically insignificant
when they would have been significant otherwise. To wit, multicollinearity is usually measured by
what are called variance inflation factors.
Once multicollinearity has been diagnosed there are several approaches to remediate it. Here
are a couple of important ones.
Principal Components Analysis. This approach casts out two or more of the original explanatory
variables and replaces them with new variables, derived from the original ones, that are
by design uncorrelated with one another. The redundancy is thus eliminated and we may
proceed as usual with the new variables in hand. Principal Components Analysis is important
for other reasons, too, not just for fixing multicollinearity problems.
Ridge Regression. The idea of this approach is to replace the original parameter estimates with a
different type of parameter estimate which is more stable under multicollinearity. The esti-
mators are not found by ordinary least squares but rather a different optimization procedure
which incorporates the variance inflation factor information.
We decided to omit a thorough discussion of multicollinearity because we are not equipped to
handle the mathematical details. Perhaps the topic will receive more attention in a later edition.
• What to do when data are not normal
o Bootstrap (see Chapter 13).
12.9.4 Akaike’s Information Criterion
AIC = -2 In L + 2(p + 1)
12.9. ADDITIONAL TOPICS
317
Chapter Exercises
Exercise 12.1. Use Equations 12.3.1, 12.3.2, and 12.3.3 to prove the Anova Equality:
SSTO = SSE + SSR.
318
CHAPTER 12. MULTIPLE LINEAR REGRESSION
Chapter 13
Resampling Methods
Computers have changed the face of statistics. Their quick computational speed and flawless ac-
curacy, coupled with large data sets acquired by the researcher, make them indispensable for many
modern analyses. In particular, resampling methods (due in large part to Bradley Efron) have
gained prominence in the modern statistician’s repertoire. We first look at a classical problem to
get some insight why.
I have seen Statistical Computing with R by Rizzo [71] and I recommend it to those looking for
a more advanced treatment with additional topics. I believe that Monte Carlo Statistical Methods
by Robert and Casella [72] has a new edition that integrates R into the narrative.
What do I want them to know?
• basic philosophy of resampling and why it is important
• resampling for standard errors and confidence intervals
• resampling for hypothesis tests (permutation tests)
13.1 Introduction
Classical question Given a population of interest, how may we effectively learn some of its salient
features, e.g., the population’s mean? One way is through representative random sampling.
Given a random sample, we summarize the information contained therein by calculating a
reasonable statistic, e.g., the sample mean. Given a value of a statistic, how do we know
whether that value is significantly different from that which was expected? We don’t; we
look at the sampling distribution of the statistic, and we try to make probabilistic assertions
based on a confidence level or other consideration. For example, we may find ourselves
saying things like, "With 95% confidence, the true population mean is greater than zero."
Problem Unfortunately, in most cases the sampling distribution is unknown. Thus, in the past,
in efforts to say something useful, statisticians have been obligated to place some restrictive
assumptions on the underlying population. For example, if we suppose that the population
has a normal distribution, then we can say that the distribution of X is normal, too, with the
319
320
CHAPTER 13. RESAMPLING METHODS
same mean (and a smaller standard deviation). It is then easy to draw conclusions, make
inferences, and go on about our business.
Alternative We don’t know what the underlying population distributions is, so let us estimate it,
just like we would with any other parameter. The statistic we use is the empirical CDF,
that is, the function that places mass 1/n at each of the observed data points (see
Section 5.5). As the sample size increases, we would expect the approximation to get better
and better (with i.i.d. observations, it does, and there is a wonderful theorem by Glivenko and
Cantelli that proves it). And now that we have an (estimated) population distribution, it is
easy to find the sampling distribution of any statistic we like: just sample from the empirical
CDF many, many times, calculate the statistic each time, and make a histogram. Done!
Of course, the number of samples needed to get a representative histogram is prohibitively
large. . . human beings are simply too slow (and clumsy) to do this tedious procedure.
Fortunately, computers are very skilled at doing simple, repetitive tasks very quickly and
accurately. So we employ them to give us a reasonable idea about the sampling distribution
of our statistic, and we use the generated sampling distribution to guide our inferences and
draw our conclusions. If we would like to have a better approximation for the sampling
distribution (within the confines of the information contained in the original sample), we
merely tell the computer to sample more. In this (restricted) sense, we are limited only by
our current computational speed and pocket book.
In short, here are some of the benefits that the advent of resampling methods has given us:
Fewer assumptions. We are no longer required to assume the population is normal or the sample
size is large (though, as before, the larger the sample the better).
Greater accuracy. Many classical methods are based on rough upper bounds or Taylor expan-
sions. The bootstrap procedures can be iterated long enough to give results accurate to sev-
eral decimal places, often beating classical approximations.
Generality. Resampling methods are easy to understand and apply to a large class of seemingly
unrelated procedures. One no longer needs to memorize long complicated formulas and
algorithms.
Remark 13.1. Due to the special structure of the empirical CDF, to get an i.i.d. sample we just need
to take a random sample of size n, with replacement, from the observed data x\, . . . ,xn. Repeats are
expected and acceptable. Since we already sampled to get the original data, the term resampling is
used to describe the procedure.
General bootstrap procedure. The above discussion leads us to the following general procedure
to approximate the sampling distribution of a statistic S = S (x\ , X2, . . . , x„) based on an observed
simple random sample x = (xi ,xi,..., x„) of size n:
1 . Create many many samples x* , . . . , x’M, called resamples, by sampling with replacement from
the data.
2. Calculate the statistic of interest S (x\), . . . , S (x*M) for each resample. The distribution of the
resample statistics is called a bootstrap distribution.
13.2. BOOTSTRAP STANDARD ERRORS
321
3. The bootstrap distribution gives information about the sampling distribution of the origi-
nal statistic S . In particular, the bootstrap distribution gives us some idea about the center,
spread, and shape of the sampling distribution of S .
13.2 Bootstrap Standard Errors
Since the bootstrap distribution gives us information about a statistic’s sampling distribution, we
can use the bootstrap distribution to estimate properties of the statistic. We will illustrate the boot-
strap procedure in the special case that the statistic S is a standard error.
Example 13.2. Standard error of the mean. In this example we illustrate the bootstrap by esti-
mating the standard error of the sample meanand we will do it in the special case that the underlying
population is norm(mean = 3, sd = 1).
Of course, we do not really need a bootstrap distribution here because from Section 8.2 we
know that X ~ norm(mean = 3, sd = 1 / sfn), but we proceed anyway to investigate how the
bootstrap performs when we know what the answer should be ahead of time.
We will take a random sample of size n = 25 from the population. Then we will resample the
data 1000 times to get 1000 resamples of size 25. We will calculate the sample mean of each of the
resamples, and will study the data distribution of the 1000 values of x.
> srs <- rnorm(25, mean = 3)
> resamps <- replicate ( 16)6)61 , sample (srs, 25, TRUE), simplify = FALSE)
> xbarstar <- sapply (resamps , mean, simplify = TRUE)
A histogram of the 1000 values of x is shown in Figure 13.2.1, and was produced by the fol-
lowing code.
> hi st (xbarstar, breaks = 4(9, prob = TRUE)
> curve (dnorm(x, 3, (9 .2), add = TRUE) # overlay true normal density
We have overlain what we know to be the true sampling distribution of X, namely, a norm(mean =
3, sd = 1/ V25) distribution. The histogram matches the true sampling distribution pretty well with
respect to shape and spread. . . but notice how the histogram is off-center a little bit. This is not a
coincidence - in fact, it can be shown that the mean of the bootstrap distribution is exactly the mean
of the original sample, that is, the value of the statistic that we originally observed. Let us calculate
the mean of the bootstrap distribution and compare it to the mean of the original sample:
> mean (xbarstar)
[1] 3.856430
> mean (srs)
[1] 3.05766
> mean (xbarstar) - mean (srs)
[1] -0.001229869
Density
322
CHAPTER 13. RESAMPLING METHODS
Histogram of xbarstar
o
2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8
xbarstar
Figure 13.2.1: Bootstrapping the standard error of the mean, simulated data
The original data were 25 observations generated from a norm(mean = 3, sd = 1) distribution. We next
resampled to get 1000 resamples, each of size 25, and calculated the sample mean for each resample. A his-
togram of the 1000 values of x is shown above. Also shown (with a solid line) is the true sampling distribution
of X , which is a norm(mean = 3, sd = 0.2) distribution. Note that the histogram is centered at the sample
mean of the original data, while the true sampling distribution is centered at the true value of /j = 3. The shape
and spread of the histogram is similar to the shape and spread of the true sampling distribution.
13.2. BOOTSTRAP STANDARD ERRORS
323
Notice how close the two values are. The difference between them is an estimate of how biased
the original statistic is, the so-called bootstrap estimate of bias. Since the estimate is so small we
would expect our original statistic (X) to have small bias, but this is no surprise to us because we
already knew from Section 8.1.1 that X is an unbiased estimator of the population mean.
Now back to our original problem, we would like to estimate the standard error of X. Looking
at the histogram, we see that the spread of the bootstrap distribution is similar to the spread of the
sampling distribution. Therefore, it stands to reason that we could estimate the standard error of X
with the sample standard deviation of the resample statistics. Let us try and see.
> sd(xbarstar)
[1] 0.2134432
We know from theory that the true standard error is 1 / V25 = 0.20. Our bootstrap estimate is
not very far from the theoretical value.
Remark 13.3. What would happen if we take more resamples? Instead of 1000 resamples, we could
increase to, say, 2000, 3000, or even 4000. . . would it help? The answer is both yes and no. Keep
in mind that with resampling methods there are two sources of randomness: that from the original
sample, and that from the subsequent resampling procedure. An increased number of resamples
would reduce the variation due to the second part, but would do nothing to reduce the variation due
to the first part. We only took an original sample of size n = 25, and resampling more and more
would never generate more information about the population than was already there. In this sense,
the statistician is limited by the information contained in the original sample.
Example 13.4. Standard error of the median. We look at one where we do not know the answer
ahead of time. This example uses the rivers data set. Recall the stemplot on page on page 42 that
we made for these data which shows them to be markedly right-skewed, so a natural estimate of
center would be the sample median. Unfortunately, its sampling distribution falls out of our reach.
We use the bootstrap to help us with this problem, and the modifications to the last example are
trivial.
> resamps <- replicate ( 16)136) , sample (rivers, 141, TRUE), simplify = FALSE)
> medstar <- sapply(resamps , median, simplify = TRUE)
> sd(medstar)
[1] 25.2366
The graph is shown in Figure 13.2.2, and was produced by the following code.
> hist(medstar, breaks = 46), prob = TRUE)
> median(rivers)
[1] 425
> mean(medstar)
[1] 426.47
> mean(medstar) - median (rivers)
[1] 1.47
Density
324
CHAPTER 13. RESAMPLING METHODS
Histogram of medstar
350 400 450 500
medstar
Figure 13.2.2: Bootstrapping the standard error of the median for the rivers data
13.2. BOOTSTRAP STANDARD ERRORS
325
Example 13.5. The boot package in R. It turns out that there are many bootstrap procedures and
commands already built into base R, in the boot package. Further, inside the boot package there
is even a function called boot. The basic syntax is of the form:
bootfdata, statistic, R)
Here, data is a vector (or matrix) containing the data to be resampled, statistic is a defined
function, of two arguments , that tells which statistic should be computed, and the parameter R
specifies how many resamples should be taken.
For the standard error of the mean (Example 13.2):
> library(boot)
> mean_fun <- function(x, indices) meanCx [indices] )
> bootCdata = srs, statistic = mean_fun, R = 100®)
ORDINARY NONPARAMETRIC BOOTSTRAP
Call:
boot(data = srs, statistic = mean_fun, R = 1880)
Bootstrap Statistics :
original bias std. error
tl* 3.85766 -8.01117741 0.2183488
For the standard error of the median (Example 13.4):
> median_fun <- function (x, indices) median (x [indices])
> boot(data = rivers, statistic = median_fun, R = 1000)
ORDINARY NONPARAMETRIC BOOTSTRAP
Call:
bootfdata = rivers, statistic = median_fun, R = 1888)
Bootstrap Statistics :
original bias std. error
tl* 425 2.003 26.87638
We notice that the output from both methods of estimating the standard errors produced similar
results. In fact, the boot procedure is to be preferred since it invisibly returns much more informa-
tion (which we will use later) than our naive script and it is much quicker in its computations.
Remark 13.6. Some things to keep in mind about the bootstrap:
• For many statistics, the bootstrap distribution closely resembles the sampling distribution
with respect to spread and shape. However, the bootstrap will not have the same center as
326
CHAPTER 13. RESAMPLING METHODS
the true sampling distribution. While the sampling distribution is centered at the population
mean (plus any bias), the bootstrap distribution is centered at the original value of the statistic
(plus any bias). The boot function gives an empirical estimate of the bias of the statistic as
part of its output.
• We tried to estimate the standard error, but we could have (in principle) tried to estimate
something else. Note from the previous remark, however, that it would be useless to estimate
the population mean /./ using the bootstrap since the mean of the bootstrap distribution is the
observed x.
• You don’t get something from nothing. We have seen that we can take a random sample
from a population and use bootstrap methods to get a very good idea about standard errors,
bias, and the like. However, one must not get lured into believing that by doing some random
resampling somehow one gets more information about the parameters than that which was
contained in the original sample. Indeed, there is some uncertainty about the parameter due
to the randomness of the original sample, and there is even more uncertainty introduced by
resampling. One should think of the bootstrap as just another estimation method, nothing
more, nothing less.
13.3 Bootstrap Confidence Intervals
13.3.1 Percentile Confidence Intervals
As a first try, we want to obtain a 95% confidence interval for a parameter. Typically the statistic
we use to estimate the parameter is centered at (or at least close by) the parameter; in such cases a
95% confidence interval for the parameter is nothing more than a 95% confidence interval for the
statistic. And to find a 95% confidence interval for the statistic we need only go to its sampling
distribution to find an interval that contains 95% of the area. (The most popular choice is the
equal-tailed interval with 2.5% in each tail.)
This is incredibly easy to accomplish with the bootstrap. We need only to take a bunch of
bootstrap resamples, order them, and choose the cr/2th and (1 -Q')th percentiles. There is a function
boot . ci in R already created to do just this. Note that in order to use the function boot . ci we
must first run the boot function and save the output in a variable, for example, data. boot. We
then plug data . boot into the function boot . ci.
Example 13.7. Percentile interval for the expected value of the median. Wee will try the naive
approach where we generate the resamples and calculate the percentile interval by hand.
> btsamps <- replicate (2GHSIS) , sample (stack, loss , 21, TRUE),
+ simplify = FALSE)
> thetast <- sapply (btsamps, median, simplify = TRUE)
> mean(thetast)
[1] 14.78
> median(stack. loss)
[1] 15
> quantile (thetast, c(Q.Q25, Q.975))
13.3. BOOTSTRAP CONFIDENCE INTERVALS
327
2.5% 97.5%
12 18
Example 13.8. Confidence interval for expected value of the median, 2nd try. Now we will do
it the right way with the boot function.
> library(boot)
> med_fun <- function(x, ind) median(x[ind])
> med_boot <- boot(stack . loss , med_fun, R = 206)6))
> boot . ci (medjooot , type = c("perc", "norm", "bca"))
BOOTSTRAP CONFIDENCE INTERVAL CALCULATIONS
Based on 208® bootstrap replicates
CALL :
boot. ci (boot. out = med_boot, type = c("perc", "norm", "bca"))
Intervals :
Level Normal Percentile BCa
95% (11.93, 18.53 ) (12.00, 18.00 ) (11.00, 18.00 )
Calculations and Intervals on Original Scale
13.3.2 Student’s t intervals (“normal intervals”)
The idea is to use confidence intervals that we already know and let the bootstrap help us when we
get into trouble. We know that a 100(1 - a)% confidence interval for the mean of a SRS ( n ) from a
normal distribution is
Z±ta/2(df = n-l)-^=, (13.3.1)
yn
where t„/2(df = n - 1) is the appropriate critical value from Student’s t distribution, and we re-
member that an estimate for the standard error of X is 5/ yfn. Of course, the estimate for the
standard error will change when the underlying population distribution is not normal, or when we
use a statistic more complicated than X. In those situations the bootstrap will give us quite rea-
sonable estimates for the standard error. And as long as the sampling distribution of our statistic is
approximately bell-shaped with small bias, the interval
statistic ± t„./2(df — n— 1) * SE(statistic) (13.3.2)
will have approximately 100(1 - a)% confidence of containing E(statistic).
Example 13.9. We will use the t-interval method to find the bootstrap CI for the median. We
have looked at the bootstrap distribution; it appears to be symmetric and approximately mound
shaped. Further, we may check that the bias is approximately 40, which on the scale of these data
is practically negligible. Thus, we may consider looking at the /-intervals. Note that, since our
sample is so large, instead of /-intervals we will essentially be using ’-intervals.
328
CHAPTER 13. RESAMPLING METHODS
Please see the handout, “Bootstrapping Confidence Intervals for the Median, 3rd try.”
We see that, considering the scale of the data, the confidence intervals compare with each other
quite well.
Remark 13.10. We have seen two methods for bootstrapping confidence intervals for a statistic.
Which method should we use? If the bias of the bootstrap distribution is small and if the distri-
bution is close to normal, then the percentile and f- intervals will closely agree. If the intervals are
noticeably different, then it should be considered evidence that the normality and bias conditions
are not met. In this case, neither interval should be used.
• BCa : bias-corrected and accelerated
o transformation invariant
o more correct and accurate
o not monotone in coverage level?
• t - intervals
o more natural
o numerically unstable
• Can do things like transform scales, compute confidence intervals, and then transform back.
• Studentized bootstrap confidence intervals where is the Studentized version of is the order
statistic of the simulation
13.4 Resampling in Hypothesis Tests
The classical two-sample problem can be stated as follows: given two groups of interest, we would
like to know whether these two groups are significantly different from one another or whether the
groups are reasonably similar. The standard way to decide is to
1. Go collect some information from the two groups and calculate an associated statistic, for
example, X\ - AS.
2. Suppose that there is no difference in the groups, and find the distribution of the statistic in
1.
3. Locate the observed value of the statistic with respect to the distribution found in 2. A value
in the main body of the distribution is not spectacular, it could reasonably have occurred by
chance. A value in the tail of the distribution is unlikely, and hence provides evidence against
the null hypothesis that the population distributions are the same.
Of course, we usually compute a p- value, defined to be the probability of the observed value of the
statistic or more extreme when the null hypothesis is true. Small /;- values are evidence against the
null hypothesis. It is not immediately obvious how to use resampling methods here, so we discuss
an example.
13.4. RESAMPLING IN HYPOTHESIS TESTS
329
Example 13.11. A study concerned differing dosages of the antiretroviral drug AZT. The common
dosage is 300mg daily. Higher doses cause more side affects, but are they significantly higher? We
examine for a 600mg dose. The data are as follows: We compare the scores from the two groups
by computing the difference in their sample means. The 300mg data were entered in xl and the
600mg data were entered into x2. The observed difference was
300 mg 284 279 289 292 287 295 285 279 306 298
600 mg 298 307 297 279 291 335 299 300 306 291
The average amounts can be found:
> mean(xl)
[1] 289.4
> mean(x2)
[1] 300.3
with an observed difference of mean(x2) - mean(xl) = 10.9. As expected, the 600 mg
measurements seem to have a higher average, and we might be interested in trying to decide if
the average amounts are significantly different. The null hypothesis should be that there is no
difference in the amounts, that is, the groups are more or less the same. If the null hypothesis
were true, then the two groups would indeed be the same, or just one big group. In that case, the
observed difference in the sample means just reflects the random assignment into the arbitrary xl
and x2 categories. It is now clear how we may resample, consistent with the null hypothesis.
Procedure:
1. Randomly resample 10 scores from the combined scores of xl and x2, and assign then to the
“xl” group. The rest will then be in the “x2” group. Calculate the difference in (re)sampled
means, and store that value.
2. Repeat this procedure many, many times and draw a histogram of the resampled statistics,
called the permutation distribution. Locate the observed difference 10.9 on the histogram to
get the p- value. If the p- value is small, then we consider that evidence against the hypothesis
that the groups are the same.
Remark 13.12. In calculating the permutation test p-value, the formula is essentially the propor-
tion of resample statistics that are greater than or equal to the observed value. Of course, this is
merely an estimate of the true p-value. As it turns out, an adjustment of +1 to both the numerator
and denominator of the proportion improves the performance of the estimated p-value, and this
adjustment is implemented in the ts . perm function.
> library (coin)
> oneway_test(len ~ supp, data = ToothGrowth )
Asymptotic 2-Sample Permutation Test
data: len by supp (OJ , VC)
Z = 1.8734, p-value = 0.061Q2
alternative hypothesis: true mu is not equal to ®
330
CHAPTER 13. RESAMPLING METHODS
13.4.1 Comparison with the Two Sample t test
We know from Chapter 10 to use the two-sample f-test to tell whether there is an improvement as a
result of taking the intervention class. Note that the f-test assumes normal underlying populations,
with unknown variance, and small sample n - 10. What does the f-test say? Below is the output.
> t . test(len ~ supp, data = ToothGrowth, alt = "greater" , var. equal = TRUE)
Two Sample t-test
data: len by supp
t = 1.9153, df = 58, p-value = 8.0302®
alternative hypothesis: true difference in means is greater than 0
95 percent confidence interval:
0.4708204 Inf
sample estimates:
mean in group OJ mean in group VC
20.66333 16.96333
The p-value for the f-test was 0.03, while the permutation test p-value was 0.061. Note that
there is an underlying normality assumption for the f-test, which isn’t present in the permutation
test. If the normality assumption may be questionable, then the permutation test would be more
reasonable. We see what can happen when using a test in a situation where the assumptions are
not met: smaller p-values. In situations where the normality assumptions are not met, for example,
small sample scenarios, the permutation test is to be preferred. In particular, if accuracy is very
important then we should use the permutation test.
Remark 13.13. Here are some things about permutation tests to keep in mind.
• While the permutation test does not require normality of the populations (as contrasted with
the f-test), nevertheless it still requires that the two groups are exchangeable; see Section 7.5.
In particular, this means that they must be identically distributed under the null hypothesis.
They must have not only the same means, but they must also have the same spread, shape,
and everything else. This assumption may or may not be true in a given example, but it will
rarely cause the f-test to outperform the permutation test, because even if the sample standard
deviations are markedly different it does not mean that the population standard deviations are
different. In many situations the permutation test will also carry over to the f-test.
• If the distribution of the groups is close to normal, then the f-test p-value and the bootstrap
p-value will be approximately equal. If they differ markedly, then this should be considered
evidence that the normality assumptions do not hold.
• The generality of the permutation test is such that one can use all kinds of statistics to com-
pare the two groups. One could compare the difference in variances or the difference in (just
about anything). Alternatively, one could compare the ratio of sample means, X \ / X2. Of
course, under the null hypothesis this last quantity should be near 1 .
• Just as with the bootstrap, the answer we get is subject to variability due to the inherent
randomness of resampling from the data. We can make the variability as small as we like by
taking sufficiently many resamples. How many? If the conclusion is very important (that is,
13.4. RESAMPLING IN HYPOTHESIS TESTS
331
if lots of money is at stake), then take thousands. For point estimation problems typically,
R = 1000 resamples, or so, is enough. In general, if the true /;- value is p then the standard
error of the estimated p-value is ^p( 1 - p)/R. You can choose R to get whatever accuracy
desired.
• Other possible testing designs:
o Matched Pairs Designs,
o Relationship between two variables.
332
CHAPTER 13. RESAMPLING METHODS
Chapter Exercises
Chapter 14
Categorical Data Analysis
This chapter is still under substantial revision. At any time you can preview any released drafts
with the development version of the IPSUR package which is available from R-Forge:
> install .packages ("IPSUR" , repos = "http://R-Forge.R-project.org")
> library (IPSUR)
> read (IPSUR)
333
334
CHAPTER 14. CATEGORICAL DATA ANALYSIS
Chapter 15
Nonparametric Statistics
This chapter is still under substantial revision. At any time you can preview any released drafts
with the development version of the IPSUR package which is available from R-Forge:
> install .packages ("IPSUR" , repos = "http://R-Forge.R-project.org")
> library (IPSUR)
> read (IPSUR)
335
336
CHAPTER 15. NONPARAMETRIC STATISTICS
Chapter 16
Time Series
This chapter is still under substantial revision. At any time you can preview any released drafts
with the development version of the IPSUR package which is available from R-Forge:
> install .packages ("IPSUR" , repos = "http://R-Forge.R-project.org")
> library (IPSUR)
> read (IPSUR)
337
338
CHAPTER 16. TIME SERIES
Appendix A
R Session Information
If you ever write the R help mailing list with a question, then you should include your session
information in the email; it makes the reader’s job easier and is requested by the Posting Guide.
Here is how to do that, and below is what the output looks like.
> sessionlnfoO
R version 2.12.2 Patched (2011-03-18 r54866)
Platform: x86_64-unknown-linux-gnu (64-bit)
locale :
[1] LC_CTYPE=en_US .UTF-8
[ 3 ] LC_TIME=en_US .UTF-8
[5] LC_MONETARY=C
[7] LC_PAPER=en_US . UTF-8
[9] LC_ADDRESS=C
[11] LC_MEASUREMENT=en_US .
LC_NUMERIC=C
LC_COLLATE=C
LC_MESSAGES=en_US .UTF-8
LC_NAME=C
LC_TELEPHONE=C
-8 LC_IDENTIFICATION=C
attached base packages:
[1] grid stats4 splines tcltk stats
[8] utils datasets methods base
graphics grDevices
other attached packages:
[1] coin_1.0-18
[3] boot_1.2-43
[5] lmtest_0 . 9-27
[7] reshape_0.8.4
[9] Hmisc_3.8-3
[11] leaps_2.9
[13] TeachingDemos_2 . 7
[15] distrEx_2.3
[17] evd_2.2-4
[19] SweaveListingUtilsj
modeltools_0 .2-17
scatterplot3d_0. 3-32
zoo_l . 6-4
plyr_l . 4
HH_2. 1-32
multcomp_l .2-5
mvtnorm_0.9-96
actuar_l .1-1
distr_2 .3.1
.5 sfsmisc_1.0-14
339
340
APPENDIX A. R SESSION INFORMATION
[21] startupmsg_0 . 7. 1 combinat_® . 0-8
[23] prob_0.9-2 diagram_l . 5 . 2
[25] shape_1.3.1 lattice_0. 19-17
[27] el071_l . 5-25 class_7.3-3
[29] qcc_2.0.1 aplpack_l . 2 . 3
[31] RcmdrPlugin.IPSUR_0. 1-7 Rcmdr_1.6-3
[33] car_2.0-9 survival_2 . 36-5
[35] nnet_7 .3-1 MASS_7.3-11
loaded via a namespace (and not attached) :
[1] cluster_l . 13 . 3 tools_2.12.2
Appendix B
GNU Free Documentation License
Version 1.3, 3 November 2008
Copyright (C) 2000, 2001, 2002, 2007, 2008 Free Software Foundation, Inc.
http://fsf.org/
Everyone is permitted to copy and distribute verbatim copies of this license document, but changing
it is not allowed.
0. PREAMBLE
The purpose of this License is to make a manual, textbook, or other functional and useful document
"free" in the sense of freedom: to assure everyone the effective freedom to copy and redistribute it,
with or without modifying it, either commercially or noncommercially. Secondarily, this License
preserves for the author and publisher a way to get credit for their work, while not being considered
responsible for modifications made by others.
This License is a kind of "copyleft", which means that derivative works of the document must
themselves be free in the same sense. It complements the GNU General Public License, which is a
copyleft license designed for free software.
We have designed this License in order to use it for manuals for free software, because free
software needs free documentation: a free program should come with manuals providing the same
freedoms that the software does. But this License is not limited to software manuals; it can be used
for any textual work, regardless of subject matter or whether it is published as a printed book. We
recommend this License principally for works whose purpose is instruction or reference.
1. APPLICABILITY AND DELINITIONS
This License applies to any manual or other work, in any medium, that contains a notice placed
by the copyright holder saying it can be distributed under the terms of this License. Such a notice
341
342
APPENDIX B. GNU FREE DOCUMENTATION LICENSE
grants a world-wide, royalty-free license, unlimited in duration, to use that work under the condi-
tions stated herein. The "Document", below, refers to any such manual or work. Any member of
the public is a licensee, and is addressed as "you". You accept the license if you copy, modify or
distribute the work in a way requiring permission under copyright law.
A "Modified Version" of the Document means any work containing the Document or a portion
of it, either copied verbatim, or with modifications and/or translated into another language.
A "Secondary Section" is a named appendix or a front-matter section of the Document that deals
exclusively with the relationship of the publishers or authors of the Document to the Document’s
overall subject (or to related matters) and contains nothing that could fall directly within that overall
subject. (Thus, if the Document is in part a textbook of mathematics, a Secondary Section may not
explain any mathematics.) The relationship could be a matter of historical connection with the
subject or with related matters, or of legal, commercial, philosophical, ethical or political position
regarding them.
The "Invariant Sections" are certain Secondary Sections whose titles are designated, as being
those of Invariant Sections, in the notice that says that the Document is released under this License.
If a section does not fit the above definition of Secondary then it is not allowed to be designated as
Invariant. The Document may contain zero Invariant Sections. If the Document does not identify
any Invariant Sections then there are none.
The "Cover Texts" are certain short passages of text that are listed, as Front-Cover Texts or
Back-Cover Texts, in the notice that says that the Document is released under this License. A
Front-Cover Text may be at most 5 words, and a Back-Cover Text may be at most 25 words.
A "Transparent" copy of the Document means a machine-readable copy, represented in a for-
mat whose specification is available to the general public, that is suitable for revising the document
straightforwardly with generic text editors or (for images composed of pixels) generic paint pro-
grams or (for drawings) some widely available drawing editor, and that is suitable for input to text
formatters or for automatic translation to a variety of formats suitable for input to text formatters.
A copy made in an otherwise Transparent file format whose markup, or absence of markup, has
been arranged to thwart or discourage subsequent modification by readers is not Transparent. An
image format is not Transparent if used for any substantial amount of text. A copy that is not
"Transparent" is called "Opaque".
Examples of suitable formats for Transparent copies include plain ASCII without markup,
Texinfo input format, ETpX input format, SGML or XML using a publicly available DTD, and
standard-conforming simple HTML, PostScript or PDF designed for human modification. Exam-
ples of transparent image formats include PNG, XCF and JPG. Opaque formats include proprietary
formats that can be read and edited only by proprietary word processors, SGML or XML for which
the DTD and/or processing tools are not generally available, and the machine-generated HTML,
PostScript or PDF produced by some word processors for output purposes only.
The "Title Page" means, for a printed book, the title page itself, plus such following pages as
are needed to hold, legibly, the material this License requires to appear in the title page. For works
in formats which do not have any title page as such, "Title Page" means the text near the most
prominent appearance of the work’s title, preceding the beginning of the body of the text.
The "publisher" means any person or entity that distributes copies of the Document to the
public.
A section "Entitled XYZ" means a named subunit of the Document whose title either is pre-
cisely XYZ or contains XYZ in parentheses following text that translates XYZ in another language.
(Here XYZ stands for a specific section name mentioned below, such as "Acknowledgements",
343
"Dedications", "Endorsements", or "History".) To "Preserve the Title" of such a section when you
modify the Document means that it remains a section "Entitled XYZ" according to this definition.
The Document may include Warranty Disclaimers next to the notice which states that this
License applies to the Document. These Warranty Disclaimers are considered to be included by
reference in this License, but only as regards disclaiming warranties: any other implication that
these Warranty Disclaimers may have is void and has no effect on the meaning of this License.
2. VERBATIM COPYING
You may copy and distribute the Document in any medium, either commercially or noncommer-
cially, provided that this License, the copyright notices, and the license notice saying this License
applies to the Document are reproduced in all copies, and that you add no other conditions whatso-
ever to those of this License. You may not use technical measures to obstruct or control the reading
or further copying of the copies you make or distribute. However, you may accept compensation
in exchange for copies. If you distribute a large enough number of copies you must also follow the
conditions in section 3.
You may also lend copies, under the same conditions stated above, and you may publicly dis-
play copies.
3. COPYING IN QUANTITY
If you publish printed copies (or copies in media that commonly have printed covers) of the Doc-
ument, numbering more than 100, and the Document’s license notice requires Cover Texts, you
must enclose the copies in covers that carry, clearly and legibly, all these Cover Texts: Lront-Cover
Texts on the front cover, and Back-Cover Texts on the back cover. Both covers must also clearly
and legibly identify you as the publisher of these copies. The front cover must present the full title
with all words of the title equally prominent and visible. You may add other material on the covers
in addition. Copying with changes limited to the covers, as long as they preserve the title of the
Document and satisfy these conditions, can be treated as verbatim copying in other respects.
If the required texts for either cover are too voluminous to fit legibly, you should put the first
ones listed (as many as fit reasonably) on the actual cover, and continue the rest onto adjacent
pages.
If you publish or distribute Opaque copies of the Document numbering more than 100, you must
either include a machine-readable Transparent copy along with each Opaque copy, or state in or
with each Opaque copy a computer-network location from which the general network-using public
has access to download using public-standard network protocols a complete Transparent copy of the
Document, free of added material. If you use the latter option, you must take reasonably prudent
steps, when you begin distribution of Opaque copies in quantity, to ensure that this Transparent
copy will remain thus accessible at the stated location until at least one year after the last time you
distribute an Opaque copy (directly or through your agents or retailers) of that edition to the public.
It is requested, but not required, that you contact the authors of the Document well before
redistributing any large number of copies, to give them a chance to provide you with an updated
version of the Document.
344
APPENDIX B. GNU FREE DOCUMENTATION LICENSE
4. MODIFICATIONS
You may copy and distribute a Modified Version of the Document under the conditions of sections
2 and 3 above, provided that you release the Modified Version under precisely this License, with
the Modified Version filling the role of the Document, thus licensing distribution and modification
of the Modified Version to whoever possesses a copy of it. In addition, you must do these things in
the Modified Version:
A. Use in the Title Page (and on the covers, if any) a title distinct from that of the Document,
and from those of previous versions (which should, if there were any, be listed in the History section
of the Document). You may use the same title as a previous version if the original publisher of that
version gives permission.
B. List on the Title Page, as authors, one or more persons or entities responsible for authorship
of the modifications in the Modified Version, together with at least five of the principal authors of
the Document (all of its principal authors, if it has fewer than five), unless they release you from
this requirement.
C. State on the Title page the name of the publisher of the Modified Version, as the publisher.
D. Preserve all the copyright notices of the Document.
E. Add an appropriate copyright notice for your modifications adjacent to the other copyright
notices.
F. Include, immediately after the copyright notices, a license notice giving the public permission
to use the Modified Version under the terms of this License, in the form shown in the Addendum
below.
G. Preserve in that license notice the full lists of Invariant Sections and required Cover Texts
given in the Document’s license notice.
H. Include an unaltered copy of this License.
I. Preserve the section Entitled "History", Preserve its Title, and add to it an item stating at least
the title, year, new authors, and publisher of the Modified Version as given on the Title Page. If
there is no section Entitled "History" in the Document, create one stating the title, year, authors,
and publisher of the Document as given on its Title Page, then add an item describing the Modified
Version as stated in the previous sentence.
J. Preserve the network location, if any, given in the Document for public access to a Transpar-
ent copy of the Document, and likewise the network locations given in the Document for previous
versions it was based on. These may be placed in the "History" section. You may omit a network
location for a work that was published at least four years before the Document itself, or if the
original publisher of the version it refers to gives permission.
K. For any section Entitled "Acknowledgements" or "Dedications", Preserve the Title of the
section, and preserve in the section all the substance and tone of each of the contributor acknowl-
edgements and/or dedications given therein.
L. Preserve all the Invariant Sections of the Document, unaltered in their text and in their titles.
Section numbers or the equivalent are not considered part of the section titles.
M. Delete any section Entitled "Endorsements". Such a section may not be included in the
Modified Version.
N. Do not retitle any existing section to be Entitled "Endorsements" or to conflict in title with
any Invariant Section.
O. Preserve any Warranty Disclaimers.
345
If the Modified Version includes new front-matter sections or appendices that qualify as Sec-
ondary Sections and contain no material copied from the Document, you may at your option des-
ignate some or all of these sections as invariant. To do this, add their titles to the list of Invariant
Sections in the Modified Version’s license notice. These titles must be distinct from any other
section titles.
You may add a section Entitled "Endorsements", provided it contains nothing but endorsements
of your Modified Version by various parties-for example, statements of peer review or that the text
has been approved by an organization as the authoritative definition of a standard.
You may add a passage of up to five words as a Front-Cover Text, and a passage of up to
25 words as a Back-Cover Text, to the end of the list of Cover Texts in the Modified Version.
Only one passage of Front-Cover Text and one of Back-Cover Text may be added by (or through
arrangements made by) any one entity. If the Document already includes a cover text for the same
cover, previously added by you or by arrangement made by the same entity you are acting on
behalf of, you may not add another; but you may replace the old one, on explicit permission from
the previous publisher that added the old one.
The author(s) and published s) of the Document do not by this License give permission to use
their names for publicity for or to assert or imply endorsement of any Modified Version.
5. COMBINING DOCUMENTS
You may combine the Document with other documents released under this License, under the terms
defined in section 4 above for modified versions, provided that you include in the combination all
of the Invariant Sections of all of the original documents, unmodified, and list them all as Invariant
Sections of your combined work in its license notice, and that you preserve all their Warranty
Disclaimers.
The combined work need only contain one copy of this License, and multiple identical Invariant
Sections may be replaced with a single copy. If there are multiple Invariant Sections with the same
name but different contents, make the title of each such section unique by adding at the end of
it, in parentheses, the name of the original author or publisher of that section if known, or else a
unique number. Make the same adjustment to the section titles in the list of Invariant Sections in
the license notice of the combined work.
In the combination, you must combine any sections Entitled "History" in the various original
documents, forming one section Entitled "History"; likewise combine any sections Entitled "Ac-
knowledgements", and any sections Entitled "Dedications". You must delete all sections Entitled
"Endorsements".
6. COLLECTIONS OL DOCUMENTS
You may make a collection consisting of the Document and other documents released under this Li-
cense, and replace the individual copies of this License in the various documents with a single copy
that is included in the collection, provided that you follow the rules of this License for verbatim
copying of each of the documents in all other respects.
You may extract a single document from such a collection, and distribute it individually under
this License, provided you insert a copy of this License into the extracted document, and follow
this License in all other respects regarding verbatim copying of that document.
346
APPENDIX B. GNU FREE DOCUMENTATION LICENSE
7. AGGREGATION WITH INDEPENDENT WORKS
A compilation of the Document or its derivatives with other separate and independent documents
or works, in or on a volume of a storage or distribution medium, is called an "aggregate" if the
copyright resulting from the compilation is not used to limit the legal rights of the compilation’s
users beyond what the individual works permit. When the Document is included in an aggregate,
this License does not apply to the other works in the aggregate which are not themselves derivative
works of the Document.
If the Cover Text requirement of section 3 is applicable to these copies of the Document, then
if the Document is less than one half of the entire aggregate, the Document’s Cover Texts may be
placed on covers that bracket the Document within the aggregate, or the electronic equivalent of
covers if the Document is in electronic form. Otherwise they must appear on printed covers that
bracket the whole aggregate.
8. TRANSLATION
Translation is considered a kind of modification, so you may distribute translations of the Docu-
ment under the terms of section 4. Replacing Invariant Sections with translations requires special
permission from their copyright holders, but you may include translations of some or all Invariant
Sections in addition to the original versions of these Invariant Sections. You may include a trans-
lation of this License, and all the license notices in the Document, and any Warranty Disclaimers,
provided that you also include the original English version of this License and the original versions
of those notices and disclaimers. In case of a disagreement between the translation and the original
version of this License or a notice or disclaimer, the original version will prevail.
If a section in the Document is Entitled "Acknowledgements", "Dedications", or "History", the
requirement (section 4) to Preserve its Title (section 1) will typically require changing the actual
title.
9. TERMINATION
You may not copy, modify, sublicense, or distribute the Document except as expressly provided
under this License. Any attempt otherwise to copy, modify, sublicense, or distribute it is void, and
will automatically terminate your rights under this License.
However, if you cease all violation of this License, then your license from a particular copyright
holder is reinstated (a) provisionally, unless and until the copyright holder explicitly and finally
terminates your license, and (b) permanently, if the copyright holder fails to notify you of the
violation by some reasonable means prior to 60 days after the cessation.
Moreover, your license from a particular copyright holder is reinstated permanently if the copy-
right holder notifies you of the violation by some reasonable means, this is the first time you have
received notice of violation of this License (for any work) from that copyright holder, and you cure
the violation prior to 30 days after your receipt of the notice.
Termination of your rights under this section does not terminate the licenses of parties who
have received copies or rights from you under this License. If your rights have been terminated and
not permanently reinstated, receipt of a copy of some or all of the same material does not give you
any rights to use it.
347
10. FUTURE REVISIONS OF THIS LICENSE
The Free Software Foundation may publish new, revised versions of the GNU Free Documentation
License from time to time. Such new versions will be similar in spirit to the present version, but
may differ in detail to address new problems or concerns. See http://www.gnu.org/copyleft/.
Each version of the License is given a distinguishing version number. If the Document specifies
that a particular numbered version of this License "or any later version" applies to it, you have the
option of following the terms and conditions either of that specified version or of any later version
that has been published (not as a draft) by the Free Software Foundation. If the Document does
not specify a version number of this License, you may choose any version ever published (not as a
draft) by the Free Software Foundation. If the Document specifies that a proxy can decide which
future versions of this License can be used, that proxy’s public statement of acceptance of a version
permanently authorizes you to choose that version for the Document.
11. RELICENSING
"Massive Multiauthor Collaboration Site" (or "MMC Site") means any World Wide Web server
that publishes copyrightable works and also provides prominent facilities for anybody to edit those
works. A public wiki that anybody can edit is an example of such a server. A "Massive Multiau-
thor Collaboration" (or "MMC") contained in the site means any set of copyrightable works thus
published on the MMC site.
"CC-BY-SA" means the Creative Commons Attribution-Share Alike 3.0 license published by
Creative Commons Corporation, a not-for-profit corporation with a principal place of business in
San Francisco, California, as well as future copyleft versions of that license published by that same
organization.
"Incorporate" means to publish or republish a Document, in whole or in part, as part of another
Document.
An MMC is "eligible for relicensing" if it is licensed under this License, and if all works
that were first published under this License somewhere other than this MMC, and subsequently
incorporated in whole or in part into the MMC, (1) had no cover texts or invariant sections, and (2)
were thus incorporated prior to November 1, 2008.
The operator of an MMC Site may republish an MMC contained in the site under CC-BY-SA
on the same site at any time before August 1, 2009, provided the MMC is eligible for relicensing.
ADDENDUM: How to use this License for your documents
To use this License in a document you have written, include a copy of the License in the document
and put the following copyright and license notices just after the title page:
Copyright (c) YEAR YOUR NAME. Permission is granted to copy, distribute and/or
modify this document under the terms of the GNU Free Documentation License, Ver-
sion 1.3 or any later version published by the Free Software Foundation; with no Invari-
ant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is
included in the section entitled "GNU Free Documentation License".
348
APPENDIX B. GNU FREE DOCUMENTATION LICENSE
If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, replace the "with... Texts."
line with this:
with the Invariant Sections being LIST THEIR TITLES, with the Front-Cover Texts
being LIST, and with the Back-Cover Texts being LIST.
If you have Invariant Sections without Cover Texts, or some other combination of the three, merge
those two alternatives to suit the situation.
If your document contains nontrivial examples of program code, we recommend releasing these
examples in parallel under your choice of free software license, such as the GNU General Public
License, to permit their use in free software.
Appendix C
History
Title:
Year:
Authors:
Publisher:
Introduction to Probability and Statistics Using R
2010
G. Jay Kerns
G. Jay Kerns
349
350
APPENDIX C. HISTORY
Appendix D
Data
This appendix is a reference of sorts regarding some of the data structures a statistician is likely to
encounter. We discuss their salient features and idiosyncrasies.
of elements, such as numbers, characters, or logical values, and there may be NA’s present. We
usually make vectors with the assignment operator <-.
> x <- c(3, 5, 9 )
Vectors are atomic in the sense that if you try to mix and match elements of different modes
then all elements will be coerced to the most convenient common mode.
> y <- c (3, "5", TRUE)
In the example all elements were coerced to character mode. We can test whether a given object
is a vector with is . vector and can coerce an object (if possible) to a vector with as . vector.
D.1.2 Matrices and Arrays
See the “Arrays and Matrices” section of An Introduction to R. Loosely speaking, a matrix is a
vector that has been reshaped into rectangular form, and an array is a multidimensional matrix.
Strictly speaking, it is the other way around: an array is a data vector with a dimension attribute
(dim), and a matrix is the special case of an array with only two dimensions. We can construct a
matrix with the matrix function.
> matrix(letters[l:6] , nrow = 2, ncol = 3)
D.l Data Structures
D.l.l Vectors
See the “Vectors and Assignment” section of An Introduction to R. A vector is an ordered sequence
[1,] '
[2,] ’
351
352
APPENDIX D. DATA
Notice the order of the matrix entries, which shows how the matrix is populated by default. We
can change this with the byrow argument:
> matrix(letters [1 : 6] , nrow = 2, ncol = 3, byrow = TRUE)
C, 1] [ 1 2] [ , 3]
[1,] "a" "b" "c"
[2,] "d" "e" "f"
We can test whether a given object is a matrix with is. matrix and can coerce an object (if
possible) to a matrix with as. matrix. As a final example watch what happens when we mix and
match types in the first argument:
> matrix(c(l , "2", NA, FALSE), nrow = 2, ncol = 3)
C, 1] [,2] [ , 3]
[1,] "1" NA "1"
[2,] "2" "FALSE" "2"
Notice how all of the entries were coerced to character for the final result (except NA). Also
notice how the four values were recycled to fill up the six entries of the matrix.
The standard arithmetic operations work element-wise with matrices.
> A <- matrix(l:6, 2, 3)
> B <- matrix(2 : 7 , 2, 3)
> A + B
C. l]
[ , 2]
[ > 3]
[1,] 3
7
11
[2,] 5
9
13
> A * B
[,1]
[ , 2]
[ . 3]
[1,] 2
12
3®
[2,] 6
2®
42
If you want the standard definition of matrix multiplication then use the %*% function. If we
were to try A %*%B we would get an error because the dimensions do not match correctly, but for
fun, we could transpose B to get conformable matrices. The transpose function t only works for
matrices (and data frames).
> try (A * B) # an error
[,1]
[ , 2]
[ > 3]
[1,1
2
12
3®
[2,]
6
2®
42
> A 9
'£*% t(B)
# this is alright
[,1]
[ , 2]
[1,]
44
53
[2,]
56
68
D.l. DATA STRUCTURES
353
To get the ordinary matrix inverse use the solve function:
> solve(A %*% t(B)) # input matrix must be square
[, 1] [ > 2]
[1,] 2.833333 -2.208333
[2,] -2.333333 1.833333
Arrays more general than matrices, and some functions (like transpose) do not work for the
more general array. Here is what an array looks like:
> array (LETTERS [1: 24] , dim = c(3,4,2))
, , 1
[,1]
[ > 2]
[ . 3]
[ . 4]
[1,] "A"
"D"
"G"
II j II
[2,] "B"
"E"
"H"
"K"
[3,] "C"
"p"
M j II
"L"
, , 2
[,1]
[ , 2]
[ > 3]
[ 1 4]
[1,] "M"
npii
"S"
"V"
[2,] "N"
"Q"
II >J' II
"W"
[3,] "0"
"R"
"U"
"X"
We can test with is . array and may coerce with as . array.
D.1.3 Data Frames
A data frame is a rectangular array of information with a special status in R. It is used as the
fundamental data structure by many of the modeling functions. It is like a matrix in that all of the
columns must be the same length, but it is more general than a matrix in that columns are allowed
to have different modes.
> x <- c(1.3, 5.2, 6)
> y <- letters[l : 3]
> z <- cCTRUE, FALSE, TRUE)
> A <- data. frame(x, y, z)
> A
x y z
1 1.3 a TRUE
2 5.2 b FALSE
3 6.0 c TRUE
Notice the names on the columns of A. We can change those with the names function.
354
APPENDIX D. DATA
> names(A) <- c("Fred", "Mary", "Sue")
> A
Fred Mary Sue
1 1.3 a TRUE
25.2 b FALSE
36.0 c TRUE
Basic command is data . frame. You can test with i s . data . frame and you can coerce with
as. data. frame.
D.1.4 Lists
A list is more general than a data frame.
D.1.5 Tables
The word “table” has a special meaning in R. More precisely, a contingency table is an object of
class “table” which is an array
Suppose you have a contingency table and would like to do descriptive or inferential statistics
on it. The default form of the table is usually inconvenient to use unless we are working with a
function specially tailored for tables. Here is how to transform your data to a more manageable
form, namely, the raw data used to make the table.
First, we coerce the table to a data frame with :
> A <- as . data . frame (Titanic)
> head (A)
Class
Sex
Age
Survived
Freq
1
1st
Male
Child
No
0
2
2nd
Male
Child
No
0
3
3rd
Male
Child
No
35
4
Crew
Male
Child
No
0
5
1st
Female
Child
No
0
6
2nd
Female
Child
No
0
Note that there are as many preliminary columns of A as there are dimensions to the table. The
rows of A contain every possible combination of levels from each of the dimensions. There is also
a Freq column, which shows how many observations there were at that particular combination of
levels.
The form of A is often sufficient for our purposes, but more often we need to do more work: we
would usually like to repeat each row of A exactly the number of times shown in the Freq column.
The reshape package [89] has the function untable designed for that very purpose:
> library (re shape)
> B <- with(A, untable (A, Freq))
> head(B)
D.l. DATA STRUCTURES
355
Class
Sex
Age
Survived
Freq
3
3rd
Male
Child
No
35
3.1
3rd
Male
Child
No
35
3.2
3rd
Male
Child
No
35
3.3
3rd
Male
Child
No
35
3.4
3rd
Male
Child
No
35
3.5
3rd
Male
Child
No
35
Now, this is more like it. Note that we slipped in a call to the with function, which was done
to make the call to untable more pretty; we could just as easily have done
untable (TitanicDF , A$Freq)
The only fly in the ointment is the lingering Freq column which has repeated values that do
not have any meaning any more. We could just ignore it, but it would be better to get rid of the
meaningless column so that it does not cause trouble later. While we are at it, we could clean up
the rownames, too.
> C <- B[, -5]
> rownames(C) <- l:dim(C) [1]
> head(C)
Class
Sex
Age
Survived
1
3rd
Male
Child
No
2
3rd
Male
Child
No
3
3rd
Male
Child
No
4
3rd
Male
Child
No
5
3rd
Male
Child
No
6
3rd
Male
Child
No
D.1.6 More about Tables
Suppose you want to make a table that looks like this:
There are at least two ways to do it.
• Using a matrix:
> tab <- matrix(l:6, nrow = 2, ncol = 3)
> rownames (tab) <- c(" first" , "second")
> colnames(tab) <- c("A", "B", "C")
> tab
ABC
first 135
second 246
o note that the columns are filled in consecutively by default. If you want to fill the data
in by rows then do byrow = TRUE in the matrix command.
356
APPENDIX D. DATA
o the object is a matrix
• Using a dataframe
> p <- c("milk", "tea")
> g <- cC'milk", "tea")
> catgs <- expand. gridCpoured = p, guessed = g)
> cnts <- c(3, 1, 1, 3)
> D <- cbind(catgs , count = cnts)
> xtabs(count ~ poured + guessed, data = D)
guessed
poured milk tea
milk 3 1
tea 1 3
o again, the data are filled in column-wise,
o the object is a dataframe
o if you want to store it as a table then do A <- xtabs(count ~ poured + guessed, data =
D)
D.2 Importing Data
Statistics is the study of data, so the statistician’s first step is usually to obtain data from somewhere
or another and read them into R. In this section we describe some of the most common sources of
data and how to get data from those sources into a running R session.
For more information please refer to the R Data Import/Export Manual, [68] and An Introduc-
tion to R, [85].
D.2.1 Data in Packages
There are many data sets stored in the datasets package of base R. To see a list of them all issue
the command datafpackage = "datasets"). The output is omitted here because the list is so
long. The names of the data sets are listed in the left column. Any data set in that list is already on
the search path by default, which means that a user can use it immediately without any additional
work.
There are many other data sets available in the thousands of contributed packages. To see
the data sets available in those packages that are currently loaded into memory issue the single
command data(). If you would like to see all of the data sets that are available in all packages
that are installed on your computer (but not necessarily loaded), issue the command
data(package = . packages (all . available = TRUE))
To load the data set foo in the contributed package bar issue the commands library (bar)
followed by data(foo), or just the single command
data(foo, package = "bar")
D.3. CREATING NEW DATA SETS
357
D.2.2 Text Files
Many sources of data are simple text files. The entries in the file are separated by delimeters such
as TABS (tab-delimeted), commas (comma separated values, or . csv, for short) or even just white
space (no special name). A lot of data on the Internet are stored with text files, and even if they are
not, a person can copy-paste information from a web page to a text file, save it on the computer,
and read it into R.
D.2.3 Other Software Files
Often the data set of interest is stored in some other, proprietary, format by third-party software
such as Minitab, SAS, or SPSS. The foreign package supports import/conversion from many of
these formats. Please note, however, that data sets from other software sometimes have properties
with no direct analogue in R. In those cases the conversion process may lose some information
which will need to be reentered manually from within R. See the Data Import/Export Manual.
As an example, suppose the data are stored in the SPSS file foo . sav which the user has copied
to the working directory; it can be imported with the commands
> library (foreign)
> read. spss(" foo. sav")
See ?read. spss for the available options to customize the file import. Note that the R Com-
mander will import many of the common file types with a menu driven interface.
D.2.4 Importing a Data Frame
The basic command is read, table.
D.3 Creating New Data Sets
Using c
Using scan
Using the R Commander.
D.4 Editing Data
D.4.1 Editing Data Values
D.4.2 Inserting Rows and Columns
D.4.3 Deleting Rows and Columns
D.4.4 Sorting Data
We can sort a vector with the sort function. Normally we have a data frame of several columns
(variables) and many, many rows (observations). The goal is to shuffle the rows so that they are
ordered by the values of one or more columns. This is done with the order function.
358
APPENDIX D. DATA
For example, we may sort all of the rows of the Puromycin data (in ascending order) by the
variable cone with the following:
> Tmp <- Puromycin [order (PuromycinSconc) , ]
> head (Tmp)
cone
rate
state
1
0.02
76
treated
2
0.02
47
treated
13
0.02
67
untreated
14
0.02
51
untreated
3
0.06
97
treated
4
0.06
107
treated
We can accomplish the same thing with the command
> with (Puromycin, Puromycin [order (cone) , J)
We can sort by more than one variable. To sort first by state and next by cone do
> with (Puromycin, Puromycin [order (state , cone), ])
If we would like to sort a numeric variable in descending order then we put a minus sign in
front of it.
> Tmp <- with (Puromycin, Puromycin [order (-cone) , J)
> head (Tmp)
cone rate state
11 1.1® 207 treated
12 1.10 200 treated
23 1.10 160 untreated
9 0.56 191 treated
10 0.56 201 treated
21 0.56 144 untreated
If we would like to sort by a character (or factor) in decreasing order then we can use the xtfrm
function which produces a numeric vector in the same order as the character vector.
> Tmp <- with (Puromycin, Puromycin[order(-xtfrm(state)) , ])
> head (Tmp)
cone rate state
13 0.02 67 untreated
14 0.02 51 untreated
15 0.06 84 untreated
16 0.06 86 untreated
17 0.11 98 untreated
18 0.11 115 untreated
D.5. EXPORTING DATA
359
D.5 Exporting Data
The basic function is write . table. The MASS package also has a write . matrix function.
D.6 Reshaping Data
• Aggregation
• Convert Tables to data frames and back
rbind, cbind
ab [order (ab[, 1]) ,]
complete . cases
aggregate
stack
360
APPENDIX D. DATA
Appendix E
Mathematical Machinery
This appendix houses many of the standard definitions and theorems that are used at some point
during the narrative. It is targeted for someone reading the book who forgets the precise definition
of something and would like a quick reminder of an exact statement. No proofs are given, and
the interested reader should consult a good text on Calculus (say, Stewart [80] or Apostol [4, 5]),
Linear Algebra (say, Strang [82] and Magnus [62]), Real Analysis (say, Folland [27], or Carothers
[12]), or Measure Theory (Billingsley [8], Ash [6], Resnick [70]) for details.
E.l Set Algebra
We denote sets by capital letters. A, B , C, etc. The letter S is reserved for the sample space, also
known as the universe or universal set, the set which contains all possible elements. The symbol 0
represents the empty set, the set with no elements.
Set Union, Intersection, and Difference
Given subsets A and B , we may manipulate them in an algebraic fashion. To this end, we have three
set operations at our disposal: union, intersection, and difference. Below is a table summarizing
the pertinent information about these operations.
Identities and Properties
1. A U0 = A, A n 0 = 0
Name
Denoted
Defined by elements
R syntax
Union
Intersection
Difference
Complement
AUB
AnB
A\B
Ac
in A or B or both
in both A and B
in A but not in B
in S but not in A
union(A, B)
intersect (A, B)
setdiff(A, B)
setdiff(S, A)
Table E.l: Set operations
361
362
APPENDIX E. MATHEMATICAL MACHINERY
2. AuS =S, AnS = A
3. AUAC = S,AnAc = 0
4. (Ac)c = A
5. The Commutative Property:
AU B - B U A, A C\ B - B C\ A (E.1.1)
6. The Associative Property:
(A U 5) U C = A U (5 U C), (AnB)nC = An(BnC) (E.1.2)
7. The Distributive Property:
All(BnC) = (AUB)n(AU B), An(BUC) = (An5)U(An5) (E.1.3)
8. DeMorgan’s Laws
(A U B)c =ACDBC and (A n B)c = Ac U BY
or more generally,
' \c ( \c
IJA“ =P|A£, and =\J Aca
a / a a / a
E.2 Differential and Integral Calculus
A function / of one variable is said to be one-to-one if no two distinct x values are mapped to the
same y = fix) value. To show that a function is one-to-one we can either use the horizontal line
test or we may start with the equation f(x\ ) = /fe) and use algebra to show that it implies x\ = X2-
Limits and Continuity
Definition E.l. Let / be a function defined on some open interval that contains the number a ,
except possibly at a itself. Then we say the limit of fix) as x approaches a is L , and we write
lim f(x) = L, (E.2.1)
x—*a
if for every e > 0 there exists a number 6 > 0 such that 0 < \x - u| < 6 implies | fix) — L\ < e.
Definition E.2. A function / is continuous at a number a if
lim fix) - f(a). (E.2.2)
x—>a
The function / is right-continuous at the number a if limv_,a+ fix) = fla), and left-continuous at a
if limv-ja- fix) = fla). Finally, the function / is continuous on an interval I if it is continuous at
every number in the interval.
(E.l. 4)
(E.l. 5)
E.2. DIFFERENTIAL AND INTEGRAL CALCULUS
363
Differentiation
Definition E.3. The derivative of a function f at a number a, denoted by f'(a), is
f'(a ) = lim
h->0
f(a + h) - f(a)
h
(E.2.3)
provided this limit exists.
A function is differentiable at a if f'(a) exists. It is differentiable on an open inten’al (a, b ) if it
is differentiable at every number in the interval.
Differentiation Rules
In the table that follows, / and g are differentiable functions and c is a constant.
o
II
o
MS
f x" = nxn~l
ax
(cfY = cf
*60
+1
II
"So
+1
^5
(fgY = f'g + fg'
(fY - f'g-fg'
\g) g2
Table E.2: Differentiation rules
Theorem E.4. Chain Rule: If f and g are both differentiable and F = f o g is the composite
function defined by F(x) = f[g(x)\, then F is differentiable and F'(x) = f'[g(xf\ ■ g'(x).
Useful Derivatives
4-e.x = ex
ax
^ In x = x 1
4- sin x = cos x
ax
4- cos x - - sin x
ax
4- tan x = sec2 x
ax
j-k tan-1 x = (1 + x2Yx
Table E.3: Some derivatives
Optimization
Definition E.5. A critical number of the function / is a value x* for which fix*) - 0 or for which
f'{x*) does not exist.
Theorem E.6. First Derivative Test. If f is differentiable and if x* is a critical number of f and if
f'(x) > 0 for x < x* and f'(x) < 0 for x > x* , then x* is a local maximum of f. If fix) < 0 for
x < x* and f'(x) > 0 for x > x* , then x * is a local minimum of f.
Theorem E.7. Second Derivative Test. If f is twice differentiable and if x* is a critical number of
f, then x* is a local maximum of f if f"(x*) < 0 and x* is a local minimum of f if f"(x*) > 0.
364
APPENDIX E. MATHEMATICAL MACHINERY
Integration
As it turns out, there are all sorts of things called “integrals”, each defined in its own idiosyncratic
way. There are Riemann integrals, Lebesgue integrals, variants of these called Stieltjes integrals,
Daniell integrals, Ito integrals, and the list continues. Given that this is an introductory book, we
will use the Riemannian integral with the caveat that the Riemann integral is not the integral that
will be used in more advanced study.
Definition E.8. Let / be defined on [a, b], a closed interval of the real line. For each n, divide
[i a,b ] into subintervals [x,, x,+ i], i = 0, 1, ... ,n — 1, of length Ax,- = (b - a)/n where xo = a and
x„ = b, and let x* be any points chosen from the respective subintervals. Then the definite integral
of / from a to b is defined by
r*b n~ 1
f(x) dx - lim V fix]) AXi, (E.2.4)
i=0
provided the limit exists, and in that case, we say that / is integrable from a to b.
Theorem E.9. The Fundamental Theorem of Calculus. Suppose f is continuous on [a, b]. Then
1. the function g defined by gix) — fit) dr, a < x < b, is continuous on [a,b] and differen-
tiable on ( a , b) with g'(x) = fix).
rb
2. I fix) dx = F(b)—F(a), where F is any antiderivative of f, that is, any function F satisfying
F' = f.
Change of Variables
Theorem E.10. If g is a differentiable function whose range is the interval [a, b ] and if both f and
g' are continuous on the range ofu = gix), then
f(u) du =
/[#(*)] g'i*) dx.
(E.2.5)
Useful Integrals
f x" dx = xJ,+1 /in + 1), nr- 1
f eA dx = ex
f x 1 dx = In |x|
J tan x dx = In | sec x|
J ax dx = ax / In a
f(x2 + 1)_1 dx = tan-1 x
Table E.4: Some integrals (constants of integration omitted)
Integration by Parts
u dv = uv — I v du
(E.2.6)
E.3. SEQUENCES AND SERIES
365
Theorem E.ll. L’Hopital’s Rule. Suppose f and g are differentiable and g'(x) 4- 0 near a, except
possibly at a. Suppose that the limit
fix)
Hm :
x-^a g O)
(E.2.7)
is an indeterminate form of type g or oo/oo. Then
m
lim :
x->a g(X)
lim
x—*a
fix)
g'(x) '
provided the limit on the right-hand side exists or is infinite.
(E.2.8)
Improper Integrals
If £ f(x) dx exists for every number t > a, then we define
f(x) dx = lim I f(x) dx.
f
(E.2.9)
Xoo
f(x)dx is convergent.
Otherwise, we say that the improper integral is divergent.
rb
If J fix) dx exists for every number t <b, then we define
nb r*b
I fix) dx = lim I fix) dx,
J-oo f^-°° Jt
(E.2.10)
rb
provided this limit exists as a finite number, and in that case we say that J fix) dx is convergent.
Otherwise, we say that the improper integral is divergent.
If both £ fix) dx and £ ^ fix) dx are convergent, then we define
fix) dx =
fix) dx +
fix) dx.
(E.2.11)
Xoo
fix) dx is convergent. Otherwise, we say that the improper integral is divergent.
E.3 Sequences and Series
A sequence is an ordered list of numbers, a i, c/2, c/3, . . . , a„ = (c/0”H . A sequence may be finite or
infinite. In the latter case we write a\, 02, ax,, . . . = fklfi ■ We say that the infinite sequence iaQf=t
converges to the finite limit L, and we write
lim ak — L, (E.3.1)
k—>oo
if for every e > 0 there exists an integer N > 1 such that \at- - L\ < e for all k > N. We say that the
infinite sequence iaff^ ° x diverges to +00 (or -00) if for every M > 0 there exists an integer N > 1
such that ak > M for all k> N (or c/3 < —M for all k > N).
366
APPENDIX E. MATHEMATICAL MACHINERY
Finite Series
Z*
k=l
1+2 + •••+« =
n(n + 1)
2
(E.3.2)
Z
£=1
£2 = l2 + 22 + • • • + n2 =
n(n + 1)(2« + 3)
(E.3.3)
The Binomial Series
2 |”J an~kbk = (a + bf (E.3.4)
Infinite Series
Given an infinite sequence of numbers a i, ak, a 3, . . . = let ,v„ denote the partial sum of the
first n terms:
n
s„ = y ak - a\ + a2 + ■ ■ ■ + an. (E.3.5)
k= 1
If the sequence (,y„)“=l converges to a finite number S then we say that the infinite series ak is
convergent and write
oo
J]ak = S. (E.3.6)
k= 1
Otherwise we say the infinite series is divergent .
Rules for Series
Let (ak)^L l and (bk)™=1 be infinite sequences and let c be a constant.
CO oo
y cak - c y ak
k=\ k= 1
oo oo oo
y (at- ± bk ) = y a;.. + y bk
k=l k= 1 (t=l
(E.3.7)
(E.3.8)
In both of the above the series on the left is convergent if the series on the right is (are) convergent.
E.3. SEQUENCES AND SERIES
367
The Geometric Series
oo
IX
1
1 - x’
XI < 1.
(E.3.9)
The Exponential Series
X
— = ex, — oo < x < oo.
(E.3.10)
Other Series
y (m + k-\ \k _ 1
m- 1 / (1-x)'"
XI < I-
(E.3.11)
X - = ln(l - x), |x| < 1.
Z— i ii
(E.3.12)
k= 1
= (1 + X)",
XI < l.
Taylor Series
If the function / has a power series representation at the point a with radius of convergence R > 0,
that is, if
OO
fix) = Yj ck(x - a)\ \x — a\< R, (E.3.14)
k=0
for some constants
(cr-)f 0, then must be
Q- =
/(»
£! ’
Jk = 0,1,2,...
(E.3.15)
Furthermore, the function / is differentiable on the open interval (a - R, a + R) with
f(x) = Yjkck ix-af~\ \x — a\< R, (E.3.16)
k=l
C+I
r-=o
Ck
(. x-a)k+1
k+ 1
\x- a\< R,
(E.3.17)
in which case both of the above series have radius of convergence R.
368
APPENDIX E. MATHEMATICAL MACHINERY
E.4 The Gamma Function
The Gamma function T will be defined in this book according to the formula
r(a) =
Fact E.12. Properties of the Gamma
• r(a) = (a - lin'd' - 1) for any
• m/2) - v^.
Xa~lQ~X dx.
Function:
a > 1, and so T(n)
for a >0. (E.4.1)
= (n — 1 )\for any positive integer n.
E.5 Linear Algebra
Matrices
A matrix is an ordered array of numbers or expressions; typically we write A = or A = |al;|.
If A has m rows and n columns then we write
A
mxn
a ii
«12 • •
&\n
«21
022 ' •
@2 n
Mm 1
0/»2
&mn-
(E.5. 1)
The identity matrix Inxn
diagonal:
is an n X n matrix with zeros everywhere except for 1 ’s along the main
I
nxn
1 0
0 1
0
0
(E.5. 2)
[0 0 1
and the matrix with ones everywhere is denoted Jnxn:
J
nxn
1 1
1 1
1
1
11-1
(E.5. 3)
A vector is a matrix with one of the dimensions equal to one, such as Amxi (a column vector)
or Alxn (a row vector). The zero vector 0nxi is an n X 1 matrix of zeros;
0„xi=[0 0 ••• Of. (E.5. 4)
The transpose of a matrix A = (ojj'j is the matrix AT = (a,;), which is just like A except the
rows are columns and the columns are rows. The matrix A is said to be symmetric if AT = A. Note
that (AB)t = BtAt.
E.6. MULTIVARIABLE CALCULUS
369
The trace of a square matrix A is the sum of its diagonal elements: tr(A) = an.
The inverse of a square matrix Anxn (when it exists) is the unique matrix denoted A-1 which sat-
isfies AA 1 = A 1 A = Inxn. If A-1 exists then we say A is invertible, or alternatively nonsingular.
Note that (at) = (A-1) .
Fact E.13. The inverse of the 2x2 matrix
A =
a b
S’
>
i
II
d
-b
c d
ad — be
-c
a
(E.5.5)
provided ad — be 4 0.
Determinants
Definition E.14. The determinant of a square matrix Anx„ is denoted det(A) or |A| and is defined
recursively by
n
det(A) = ^(- 1 T’uij det(M,7), (E.5.6)
i= 1
where My is the submatrix formed by deleting the ;th row and /h column of A. We may choose
any fixed 1 < j < n we wish to compute the determinant; the final result is independent of the j
chosen.
Fact E.15. The determinant of the 2x2 matrix
A =
a
c
b
d
is |A| - ad — be.
Fact E.16. A square matrix A is nonsingular if and only if det(A) 4 0.
(E.5.7)
Positive (Semi)Definite
If the matrix A satisfies xTAx > 0 for all vectors x 4 0, then we say that A is positive semidefinite.
If strict inequality holds for all x 4 0, then A is positive definite. The connection to statistics is that
covariance matrices (see Chapter 7) are always positive semidefinite, and many of them are even
positive definite.
E.6 Multivariable Calculus
Partial Derivatives
If / is a function of two variables, its first-order partial derivatives are defined by
df d
f(x + It, y) - f(x,y)
and
df
dy
d_
dy
fix, y) = lim
J h^O
fix, y + h )- fix, y)
(E.6.1)
h
(E.6. 2)
370
APPENDIX E. MATHEMATICAL MACHINERY
provided these limits exist. The second-order partial derivatives of / are defined by
92f 3 (df\ d2f d I8f\
dx2 dx \ dx J ’ dy2 dy \ dy J ’ dxdy dxydy)' dydx dy \ dx J '
In many cases (and for all cases in this book) it is true that
d2f d2f
dxdy dydx '
(E.6.3)
(E.6.4)
Optimization
An function / of two variables has a local maximum at (a, b) if f(x, y) > f (a, b) for all points (x, y)
near (a, b), that is, for all points in an open disk centered at (a, b ). The number f{a, b) is then called
a local maximum value of /. The function / has a local minimum if the same thing happens with
the inequality reversed.
Suppose the point (a, b) is a critical point of /, that is, suppose (a, b ) satisfies
df , ... df
—(a,b) = — (a, b) = 0.
ox oy
(E.6.5)
Further suppose |4 and are continuous near (a, b ). Let the Hessian matrix H (not to be con-
fused with the hat matrix H of Chapter 12) be defined by
- cPj_ &J_
H= jg- df[ . (E.6.6)
. dydx dy 2 .
We use the following rules to decide whether (a, b) is an extremum (that is, a local minimum or
local maximum) of /.
Q 2 r
• If det (H) > 0 and j^(a, b) > 0, then (a, b ) is a local minimum of /.
Q 2 f
• If det(H) > 0 and ^4(a, b) < 0, then (a, b ) is a local maximum of /.
• If det (H) < 0, then ( a , b) is a saddle point of / and so is not an extremum of /.
• If det (H) = 0, then we do not know the status of (a, b)\ it might be an extremum or it might
not be.
Double and Multiple Integrals
Let / be defined on a rectangle R = [a, h\ x [c, d], and for each m and n divide [a, h\ (respectively
[c,dY> into subintervals [x;, Xj+ 1], i = 0, 1, ... , m— 1 (respectively yi+i]) of length A Xj = ( b-d)/m
(respectively Ay, = (d - c)/n) where xq = a and xm = b (and y0 = c and yn = d), and let x* (y*)
be any points chosen from their respective subintervals. Then the double integral of / over the
rectangle R is
d b
/-> y~\ s-\ tl til
JJ f(x,y ) d,4 = J l f(x,y ) dxdy = Jirn^ ^ f(x*,y*)AxJAyl,
(E.6.7)
i=i j= l
provided this limit exists. Multiple integrals are defined in the same way just with more letters and
sums.
E.6. MULTIVARIABLE CALCULUS
371
Bivariate and Multivariate Change of Variables
Suppose we have a transformation1 T that maps points (u, v) in a set A to points (x, y) in a set B. We
typically write x = x{u, v) and y = y(u, v), and we assume that x and y have continuous first-order
partial derivatives. We say that T is one-to-one if no two distinct (m, v) pairs get mapped to the
same ( x,y ) pair; in this book, all of our multivariate transformations T are one-to-one.
The Jacobian (pronounced “yah-KOH-bee-uhn”) of T is denoted by d(x,y)/d(u,v) and is de-
fined by the determinant of the following matrix of partial derivatives:
d(x,y)
d (u, v)
dx
(lit
dy
du
dx
dv
dx dy dx dy
du dv dv du
(E.6.8)
If the function / is continuous on A and if the Jacobian of T is nonzero except perhaps on the
boundary of A, then
JJ f(x, y) dx dy = JJ f [x(u, v), y(u, v)]
d(x,y)
d{u, v)
di< dv.
(E.6.9)
A multivariate change of variables is defined in an analogous way: the one-to-one transformation
T maps points (u\,U2, ■ ■ ■ , u„) to points (x\ , X2, . . . , xn), the Jacobian is the determinant of the n x n
matrix of first-order partial derivatives of T (lined up in the natural manner), and instead of a double
integral we have a multiple integral over multidimensional sets A and B.
1 For our purposes T is in fact the inverse of a one-to-one transformation that we are initially given. We usually start
with functions that map ( x,y ) i — > (m, v), and one of our first tasks is to solve for the inverse transformation that maps
(«, v) i — > ( x,y ). It is this inverse transformation which we are calling T.
372
APPENDIX E. MATHEMATICAL MACHINERY
Appendix F
Writing Reports with R
Perhaps the most important part of a statistician’s job once the analysis is complete is to commu-
nicate the results to others. This is usually done with some type of report that is delivered to the
client, manager, or administrator. Other situations that call for reports include term papers, final
projects, thesis work, etc. This chapter is designed to pass along some tips about writing reports
once the work is completed with R.
F.l What to Write
It is possible to summarize this entire appendix with only one sentence: the statistician ’s goal is
to communicate with others. To this end, there are some general guidelines that I give to students
which are based on an outline originally written and shared with me by Dr. G. Andy Chang.
Basic Outline for a Statistical Report
1 . Executive Summary (a one page description of the study and conclusion)
2. Introduction
(a) What is the question, and why is it important?
(b) Is the study observational or experimental?
(c) What are the hypotheses of interest to the researcher?
(d) What are the types of analyses employed? (one sample f-test, paired-sample f-test,
ANOVA, chi-square test, regression, . . . )
3. Data Collection
(a) Describe how the data were collected in detail.
(b) Identify all variable types: quantitative, qualitative, ordered or nominal (with levels),
discrete, continuous.
(c) Discuss any limitations of the data collection procedure. Look carefully for any sources
of bias.
373
374
APPENDIX F. WRITING REPORTS WITH R
4. Summary Information
(a) Give numeric summaries of all variables of interest.
i. Discrete: (relative) frequencies, contingency tables, odds ratios, etc.
ii. Continuous: measures of center, spread, shape.
(b) Give visual summaries of all variables of interest.
i. Side-by-side boxplots, scatterplots, histograms, etc.
(c) Discuss any unusual features of the data (outliers, clusters, granularity, etc.)
(d) Report any missing data and identify any potential problems or bias.
5. Analysis
(a) State any hypotheses employed, and check the assumptions.
(b) Report test statistics, /;- values, and confidence intervals.
(c) Interpret the results in the context of the study.
(d) Attach (labeled) tables and/or graphs and make reference to them in the report as
needed.
6. Conclusion
(a) Summarize the results of the study. What did you learn?
(b) Discuss any limitations of the study or inferences.
(c) Discuss avenues of future research suggested by the study.
F.2 How to Write It with R
Once the decision has been made what to write, the next task is to typeset the information to be
shared. To do this the author will need to select software to use to write the documents. There are
many options available, and choosing one over another is sometimes a matter of taste. But not all
software were created equal, and R plays better with some applications than it does with others.
In short, R does great with DTpX and there are many resources available to make writing a
document with R and DTpX easier. But DTpX is not for the beginner, and there are other word
processors which may be acceptable depending on the circumstances.
F.2.1 Microsoft® Word
It is a fact of life that Microsoft® Windows is currently the most prevalent desktop operating system
on the planet. Those who own Windows also typically own some version of Microsoft Office, thus
Microsoft Word is the default word processor for many, many people.
The standard way to write an R report with Microsoft® Word is to generate material with R and
then copy-paste the material at selected places in a Word document. An advantage to this approach
is that Word is nicely designed to make it easy to copy-and-paste from RGui to the Word document.
F.2. HOW TO WRITE IT WITH R
375
A disadvantage to this approach is that the R input/output needs to be edited manually by the
author to make it readable for others. Another disadvantage is that the approach does not work on
all operating systems (not on Linux, in particular). Yet another disadvantage is that Microsoft®
Word is proprietary, and as a result, R does not communicate with Microsoft® Word as well as it
does with other software as we shall soon see.
Nevertheless, if you are going to write a report with Word there are some steps that you can
take to make the report more amenable to the reader.
1. Copy and paste graphs into the document. You can do this by right clicking on the graph and
selecting Copy as bitmap, or Copy as metafile, or one of the other options. Then move the
cursor to the document where you want the picture, right-click, and select Paste.
2. Resize (most) pictures so that they take up no more than 1/2 page. You may want to put
graphs side by side; do this by inserting a table and placing the graphs inside the cells.
3. Copy selected R input and output to the Word document. All code should be separated from
the rest of the writing, except when specifically mentioning a function or object in a sentence.
4. The font of R input/output should be Courier New, or some other monowidth font (not Times
New Roman or Calibri); the default font size of 12 is usually too big for R code and should
be reduced to, for example, lQpt.
It is also possible to communicate with R through OpenOffice.org, which can export to the propri-
etary ( . doc) format.
F.2.2 OpenOffice.org and odfWeave
OpenOffice.org (OO.o) is an open source desktop productivity suite which mirrors Microsoft® Of-
fice. It is especially nice because it works on all operating systems. OO.o can read most document
formats, and in particular, it will read . doc files. The standard OO.o file extension for documents
is . odt, which stands for “open document text”.
The odfWeave package [55] provides a way to generate an . odt file with R input and output
code formatted correctly and inserted in the correct places, without any additional work. In this
way, one does not need to worry about all of the trouble of typesetting R output. Another advantage
of odfWeave is that it allows you to generate the report dynamically; if the data underlying the
report change or are updated, then a few clicks (or commands) will generate a brand new report.
One disadvantage is that the source . odt file is not easy to read, because it is difficult to visually
distinguish the noweb parts (where the R code is) from the non-noweb parts. This can be fixed by
manually changing the font of the noweb sections to, for instance. Courier font, size l®pt. But it
is extra work. It would be nice if a program would discriminate between the two different sections
and automatically typeset the respective parts in their correct fonts. This is one of the advantages
to LyX.
Another advantage of OO.o is that even after you have generated the outfile, it is fully editable
just like any other . odt document. If there are errors or formatting problems, they can be fixed at
any time.
Here are the basic steps to typeset a statistical report with OO.o.
1 . Write your report as an . odt document in OO.o just as you would any other document. Call
this document infile . odt, and make sure that it is saved in your working directory.
376
APPENDIX F. WRITING REPORTS WITH R
2. At the places you would like to insert R code in the document, write the code chunks in the
following format:
«»=
x <- rnorm(lQ)
mean(x)
@
or write whatever code you want between the symbols «»= and
3. Open R and type the following:
> library(odfWeave)
> odfWeave(file = "infile. odt" , dest = "out file. odt")
4. The compiled ( . odt) file, complete with all of the R output automatically inserted in the
correct places, will now be the file outfile.odt located in the working directory. Open
outfile . odt, examine it, modify it, and repeat if desired.
There are all sorts of extra things that can be done. For example, the R commands can be suppressed
with the tag «echo = FALSE»=, and the R output may be hidden with «results = hide»=.
See the odfWeave package documentation for details.
F.2.3 Sweave and ETjX
This approach is nice because it works for all operating systems. One can quite literally typeset
anything with OTpX . All of this power comes at a price, however. The writer must learn the ETpX
language which is a nontrivial enterprise. Even given the language, if there is a single syntax error,
or a single delimeter missing in the entire document, then the whole thing breaks.
ETgX can do anything, but it is relatively difficult to learn and very grumpy about syntax errors
and delimiter matching, there are however programs useful for formatting ETjgX.
A disadvantage is that you cannot see the mathematical formulas until you run the whole file
with ETpX .
A disadvantage is that figures and tables are relatively difficult.
There are programs to make the process easier AUCTpX
dev.copy2eps, also dev.copy2pdf
http : / /www . stat . uni-muenchen . de/~leisch/ Sweave/
F.2.4 Sweave and LyX
This approach is nice because it works for all operating systems, ft gives you everything from the
last section and makes it easier to use ETpX. That being said, it is better to know ETpX already
when migrating to LyX, because you understand all of the machinery going on under the hood.
Program Listings and the R language
This book was written with LyX.
http : / /gregor . gor j anc . googlepages . com/lyx- sweave
F.3. FORMATTING TABLES
377
F.3 Formatting Tables
The prettyR package
the Hmisc package
> library (Hmisc)
> summary (cbind (Sepal. Length, Sepal .Width) ~ Species, data = iris)
cbind(Sepal .Length, Sepal. Width) N=150
+ +
1 1
. + + +
|N | Sepal .Length | Sepal .Width
I Species | setosa | 5®|5.®®6®Q®
| | versicolor | 5® | 5 . 93608®
| Ivirginica | 5® | 6 . 5880Q®
13.42888®
I2.778Q8®
12.97488®
| Overall |
+ +
1 15®| 5.843333
. + +
13.857333
+
There is a method argument to summary, which is set to method = "response" by default.
There are two other methods for summarizing data: reverse and cross. See ?summary . formula or
the following document from Frank Harrell for more details http : //biostat .me .Vanderbilt .
edu/twiki/bin/view/Main/StatReport.
F.4 Other Formats
HTML and prettyR
R2HTML
378
APPENDIX F. WRITING REPORTS WITH R
Appendix G
Instructions for Instructors
WARNING: this appendix is not applicable until the exercises have been written.
Probably this book could more accurately be described as software. The reason is that the
document is one big random variable, one observation realized out of millions. It is electronically
distributed under the GNU FDL, and “free” in both senses: speech and beer.
There are four components to IfsUR: the Document, the Program used to generate it, the R
package that holds the Program, and the Ancillaries that accompany it.
The majority of the data and exercises have been designed to be randomly generated. Different
realizations of this book will have different graphs and exercises throughout. The advantage of this
approach is that a teacher, say, can generate a unique version to be used in his/her class. Students
can do the exercises and the teacher will have the answers to all of the problems in their own,
unique solutions manual. Students may download a different solutions manual online somewhere
else, but none of the answers will match the teacher’s copy.
Then next semester, the teacher can generate a new book and the problems will be more or less
identical, except the numbers will be changed. This means that students from different sections of
the same class will not be able to copy from one another quite so easily. The same will be true for
similar classes at different institutions. Indeed, as long as the instructor protects his/her key used
to generate the book, it will be difficult for students to crack the code. And if they are industrious
enough at this level to find a way to (a) download and decipher my version’s source code, (b)
hack the teacher’s password somehow, and (c) generate the teacher’s book with all of the answers,
then they probably should be testing out of an “Introduction to Probability and Statistics” course,
anyway.
The book that you are reading was created with a random seed which was set at the beginning.
The original seed is 42. You can choose your own seed, and generate a new book with brand new
data for the text and exercises, complete with updated manuals. A method I recommend for finding
a seed is to look down at your watch at this very moment and record the 6 digit hour, minute, and
second (say, 9:52:59am): choose that for a seed1. This method already provides for over 43,000
books, without taking military time into account. An alternative would be to go to R and type
> options(digits = 16)
> runif(l)
'in fact, this is essentially the method used by R to select an initial random seed (see ?set.seed). However, the instructor
should set the seed manually so that the book can be regenerated at a later time, if necessary.
379
380
APPENDIX G. INSTRUCTIONS FOR INSTRUCTORS
[1] 0.2170129411388189
Now choose 2170129411388188 as your secret seed. . . write it down in a safe place and do not
share it with anyone. Next generate the book with your seed using LyX-Sweave or Sweave-LTpX.
You may wish to also generate Student and Instructor Solution Manuals. Guidance regarding this
is given below in the How to Use This Document section.
G.l Generating This Document
You will need three (3) things to generate this document for yourself, in addition to a current
R distribution which at the time of this writing is R version 2.12.2 Patched (2011-03-18
r54866) :
1 . a LTpX distribution,
2. Sweave (which comes with R automatically), and
3. LyX (optional, but recommended).
We will discuss each of these in turn.
BTjjX: The distribution used by the present author was TjtX Live (http : / /www . tug . org/ texlive/).
There are plenty of other perfectly suitable DTftX distributions depending on your operating
system, one such alternative being MikTpX (http://miktex.org/) for Microsoft Win-
dows.
Sweave: If you have R installed, then the required Sweave files are already on your system. . .
somewhere. The only problems that you may have are likely associated with making sure
that your LTpX distribution knows where to find the Sweave . sty file. See the Sweave
Homepage (http://www.statistik.lmu.de/~leisch/Sweave/) for guidance on how
to get it working on your particular operating system.
LyX: Strictly speaking, LyX is not needed to generate this document. But this document was
written stem to stern with LyX, taking full advantage of all of the bells and whistles that
LyX has to offer over plain LTpX editors. And it’s free. See the LyX homepage (http :
//www . lyx . org/) for additional information.
If you decide to give LyX a try, then you will need to complete some extra steps to coordinate
Sweave and LyX with each other. Luckily, Gregor Gorjanc has a website and an R News
article [36] to help you do exactly that. See the LyX-Sweave homepage (http : //gregor .
gor j anc . googlepages . com/lyx- sweave) for details.
An attempt was made to not be extravagant with fonts or packages so that a person would not need
the entire CTAN (or CRAN) installed on their personal computer to generate the book. Nevertheless,
there are a few extra packages required. These packages are listed in the preamble of IPSUR . Rnw,
IPSUR . tex, and IPSUR . lyx.
G.2. HOW TO USE THIS DOCUMENT
381
G.2 How to Use This Document
The easiest way to use this document is to install the IPSUR package from CRAN and be all done.
This way would be acceptable if there is another, primary, text being used for the course and FsUR
is only meant to play a supplementary role.
If you plan for IlgUR to serve as the primary text for your course, then it would be wise to
generate your own version of the document. You will need the source code for the Program which
can be downloaded from CRAN or the IRjUR website. Once the source is obtained there are four (4)
basic steps to generating your own copy.
1. Randomly select a secret “seed” of integers and replace my seed of 42 with your own seed.
2. Make sure that the maintext branch is turned ON and also make sure that both the solutions
branch and the answers branch are turned OFF. Use LyX or your DTpX editor with Sweave
to generate your unique PDF copy of the book and distribute this copy to your students. (See
the LyX User’s Guide to learn more about branches; the ones referenced above can be found
under Document > Settings > Branches.)
3. Turn the maintext branch" OFF and the solutions branch ON. Generate a “Student Solu-
tions Manual” which has complete solutions to selected exercises and distribute the PDF to
the students.
4. Leave the solutions branch ON and also turn the answers branch ON and generate an
“Instructor Solutions and Answers Manual” with full solutions to some of the exercises and
just answers to the remaining exercises. Do NOT distribute this to the students - unless of
course you want them to have the answers to all of the problems.
To make it easier for those people who do not want to use LyX (or for whatever reason cannot get it
working), I have included three (3) Sweave hies corresponding to the main text, student solutions,
and instructor answers, that are included in the IR;UR source package in the /tex subdirectory. In
principle it is possible to change the seed and generate the three parts separately with only Sweave
and DTpX. This method is not recommended by me, but is perhaps desirable for some people.
Generating Quizzes and Exams
• you can copy paste selected exercises from the text, put them together, and you have a
quiz. Since the numbers are randomly generated you do not need to worry about different
semesters. And you will have answer keys already for all of your QUIZZES and EXAMS,
too.
G.3 Ancillary Materials
In addition to the main text, student manual, and instructor manual, there is IPSUR.R which is all
of the parsed R code used in the document.
2 You can leave the maintext branch ON when generating the solutions manuals, but (1) all of the page numbers will be
different, and (2) the typeset solutions will generate and take up a lot of space between exercises.
382
APPENDIX G. INSTRUCTIONS FOR INSTRUCTORS
G.4 Modifying This Document
Since this document is released under the GNU-FDL, you are free to modify this document however
you wish (in accordance with the license - see Appendix B). The immediate benefit of this is that
you can generate the book, with brand new problem sets, and distribute it to your students simply
as a PDF (in an email, for instance). As long as you distribute less than 100 such Opaque copies,
you are not even required by the GNU-FDL to share your Transparent copy (the source code with
the secret key) that you used to generate them. Next semester, choose a new key and generate a
new copy to be distributed to the new class.
But more generally, if you are not keen on the way I explained (or failed to explain)
something, then you are free to rewrite it. If you would like to cover more (or less)
material, then you are free to add (or delete) whatever Chapters/Sections/Paragraphs
that you wish. And since you have the source code, you do not need to retype the
wheel.
Some individuals will argue that the nature of a statistics textbook like this one, many of the exer-
cises being randomly generated by design , does a disservice to the students because the exercises
do not use real-world data. That is a valid criticism. . . but in my case the benefits outweighed
the detriments and I moved forward to incorporate static data sets whenever it was feasible and
effective. Frankly, and most humbly, the only response I have for those individuals is: “Please refer
to the preceding paragraph.”
Appendix H
RcmdrTestDrive Story
The goal of RcmdrTestDrive was to have a data set sufficiently rich in the types of data represented
such that a person could load it into the R Commander and be able to explore all of Rcmdr’s menu
options at once. I decided early-on that an efficient way to do this would be to generate the data
set randomly, and later add to the list of variables as more Rcmdr menu options became available.
Generating the data was easy, but generating a story that related all of the respective variables
proved to be less so.
In the Summer of 2006 I gave a version of the raw data and variable names to my STAT 3743
Probability and Statistics class and invited each of them to write a short story linking all of the
variables together in a coherent narrative. No further direction was given.
The most colorful of those I received was written by Jeffery Cornfield, submitted July 12, 2006,
and is included below with his permission. It was edited slightly by the present author and updated
to respond dynamically to the random generation of RcmdrTestDrive; otherwise, the story has
been unchanged.
Case File: ALU-179 “Murder Madness in Toon Town”
*** WARNING***
***This file is not for the faint of heart, dear reader, because it is filled with horrible
images that will haunt your nightmares. If you are weak of stomach, have irritable
bowel syndrome, or are simply paranoid, DO NOT READ FURTHER! Otherwise,
read at your own risk.***
One fine sunny day. Police Chief R. Runner called up the forensics department at Acme-Looney
University. There had been 166 murders in the past 7 days, approximately one murder every hour,
of many of the local Human workers, shop keepers, and residents of Toon Town. These alarming
rates threatened to destroy the fragile balance of Toon and Human camaraderie that had developed
in Toon Town.
Professor Twee T. Bird, a world-renowned forensics specialist and a Czechoslovakian native,
received the call. “Professor, we need your expertise in this field to identify the pattern of the killer
383
384
APPENDIX H. RCMDRTESTDRIVE STORY
or killers,” Chief Runner exclaimed. “We need to establish a link between these people to stop this
massacre.”
“Yes, Chief Runner, please give me the details of the case,” Professor Bird declared with a
heavy native accent, (though, for the sake of the case file, reader, I have decided to leave out the
accent due to the fact that it would obviously drive you - if you will forgive the pun - looney!)
“All prints are wiped clean and there are no identifiable marks on the bodies of the victims. All
we are able to come up with is the possibility that perhaps there is some kind of alternative method
of which we are unaware. We have sent a secure e-mail with a listing of all of the victims’ races,
genders, locations of the bodies, and the sequential order in which they were killed. We have also
included other information that might be helpful,” said Chief Runner.
“Thank you very much. Perhaps I will contact my colleague in the Statistics Department here.
Dr. Elmer Fudd-Einstein,” exclaimed Professor Bird. “He might be able to identify a pattern of
attack with mathematics and statistics.”
“Good luck trying to find him. Professor. Last I heard, he had a bottle of scotch and was in
the Hundred Acre Woods hunting rabbits,” Chief Runner declared in a manner that questioned the
beloved doctor’s credibility.
“Perhaps I will take a drive to find him. The fresh air will do me good.”
***I will skip ahead, dear reader, for much occurred during this time. Needless to
say, after a fierce battle with a mountain cat that the Toon-ology Department tagged
earlier in the year as “Sylvester,” Professor Bird found Dr. Fudd-Einstein and brought
him back, with much bribery of alcohol and the promise of the future slaying of those
“wascally wabbits” (it would help to explain that Dr. Fudd-Einstein had a speech im-
pediment which was only worsened during the consumption of alcohol.)***
Once our two heroes returned to the beautiful Acme-Looney University, and once Dr. Fudd-Einstein
became sober and coherent, they set off to examine the case and begin solving these mysterious
murders.
“First off,” Dr. Fudd-Einstein explained, “these people all worked at the University at some
point or another. Also, there also seems to be a trend in the fact that they all had a salary between
$12 and $21 when they retired.”
“That’s not really a lot to live off of,” explained Professor Bird.
“Yes, but you forget that the Looney Currency System works differently than the rest of the
American Currency System. One Looney is equivalent to Ten American Dollars. Also, these
faculty members are the ones who faced a cut in their salary, as denoted by ‘reduction’. Some of
them dropped quite substantially when the University had to fix that little faux pas in the Chemistry
Department. You remember: when Dr. D. Duck tried to create that ‘Everlasting Elixir?’ As a
result, these faculty left the university. Speaking of which, when is his memorial service?” inquired
Dr. Fudd-Einstein.
“This coming Monday. But if there were all of these killings, how in the world could one person
do it? It just doesn’t seem to be possible; stay up 7 days straight and be able to kill all of these
people and have the energy to continue on,” Professor Bird exclaimed, doubting the guilt of only
one person.
“Perhaps then, it was a group of people, perhaps there was more than one killer placed through-
out Toon Town to commit these crimes. If I feed in these variables, along with any others that
might have a pattern, the Acme Computer will give us an accurate reading of suspects, with a
scant probability of error. As you know, the Acme Computer was developed entirely in house
385
here at Acme-Looney University,” Dr. Fudd-Einstein said as he began feeding the numbers into the
massive server.
“Hey, look at this,” Professor Bird exclaimed, “What’s with this before/after information?”
“Scroll down; it shows it as a note from the coroner’s office. Apparently Toon Town Coroner
Marvin - that strange fellow from Mars, Pennsylvania - feels, in his opinion, that given the fact
that the cadavers were either smokers or non-smokers, and given their personal health, and family
medical history, that this was their life expectancy before contact with cigarettes or second-hand
smoke and after,” Dr. Fudd-Einstein declared matter-of-factly.
“Well, would race or gender have something to do with it, Elmer?” inquired Professor Bird.
“Maybe, but I would bet my money on somebody was trying to quiet these faculty before they
made a big ruckus about the secret money-laundering of Old Man Acme. You know, most people
think that is how the University receives most of its funds, through the mob families out of Chicago.
And I would be willing to bet that these faculty figured out the connection and were ready to tell
the Looney Police.” Dr. Fudd-Einstein spoke lower, fearing that somebody would overhear their
conversation.
Dr. Fudd-Einstein then pressed Enter on the keyboard and waited for the results. The massive
computer roared to life. . . and when I say roared, I mean it literally roared. All the hidden bells,
whistles, and alarm clocks in its secret compartments came out and created such a loud racket that
classes across the university had to come to a stand-still until it finished computing.
Once it was completed, the computer listed 4 names:
Yosemite Sam (“Looney” Insane Asylum)
Wile E. Coyote (deceased)
Foghorn Leghorn (whereabouts unknown)
Granny (1313 Mockingbird Lane, Toon Town USA)
Dr. Fudd-Einstein and Professor Bird looked on in silence. They could not believe their eyes. The
greatest computer on the Gulf of Mexico seaboard just released the most obscure results imagin-
able.
“There seems to be a mistake. Perhaps something is off,” Professor Bird asked, still unable to
believe the results.
“Not possible; the Acme Computer takes into account every kind of connection available. It
considers affiliations to groups, and affiliations those groups have to other groups. It checks the
FBI, CIA, British intelligence, NAACP, AARP, NSA, JAG, TWA, EPA, FDA, USWA, R, MAPLE,
SPSS, SAS, and Ben & Jerry’s files to identify possible links, creating the most powerful computer
in the world. . . with a tweak of Toon fanaticism,” Dr. Fudd-Einstein proclaimed, being a proud
co-founder of the Acme Computer Technology.
“Wait a minute, Ben & Jerry? What would eating ice cream have to do with anything?” Profes-
sor Bird inquired.
“It is in the works now, but a few of my fellow statistician colleagues are trying to find a
mathematical model to link the type of ice cream consumed to the type of person they might
become. Assassins always ate vanilla with chocolate sprinkles, a little known fact they would tell
you about Oswald and Booth,” Dr. Fudd-Einstein declared.
386
APPENDIX H. RCMDRTESTDRIVE STORY
“I’ve heard about this. My forensics graduate students are trying to identify car thieves with ei-
ther rocky road or mint chocolate chip. . . so far, the pattern is showing a clear trend with chocolate
chip,” Professor Bird declared.
“Well, what do we know about these suspects. Twee?” Dr. Fudd-Einstein asked.
“Yosemite Sam was locked up after trying to rob that bank in the West Borough. Apparently
his guns were switched and he was sent the Acme Kids Joke Gun and they blew up in his face. The
containers of peroxide they contained turned all of his facial hair red. Some little child is running
around Toon Town with a pair of .38’s to this day.
“Wile E. Coyote was that psychopath working for the Yahtzee - the fanatics who believed that
Toons were superior to Humans. He strapped sticks of Acme Dynamite to his chest to be a martyr
for the cause, but before he got to the middle of Toon Town, this defective TNT blew him up. Not
a single other person - Toon or Human - was even close.
“Foghorn Leghorn is the most infamous Dog Kidnapper of all times. He goes to the homes
of prominent Dog citizens and holds one of their relatives for ransom. If they refuse to pay, he
sends them to the pound. Either way, they’re sure stuck in the dog house,” Professor Bird laughed.
Dr. Fudd-Einstein didn’t seem amused, so Professor Bird continued.
“Granny is the most beloved alumnus of Acme-Looney University. She was in the first graduat-
ing class and gives graciously each year to the university. Without her continued financial support,
we wouldn’t have the jobs we do. She worked as a parking attendant at the University lots. . . wait
a minute, take a look at this,” Professor Bird said as he scrolled down in the police information.
“Granny’s signature is on each of these faculty members’ parking tickets. Kind of odd, consider-
ing the Chief-of-Parking signed each personally. The deceased had from as few as 1 ticket to as
many as 18. All tickets were unpaid.
“And look at this. Granny married Old Man Acme after graduation. He was a resident of
Chicago and rumored to be a consigliere to one of the most prominent crime families in Chicago,
the Chuck Jones/Warner Crime Family,” Professor Bird read from the screen as a cold feeling of
terror rose from the pit of his stomach.
“Say, don’t you live at her house? Wow, you’re living under the same roof as one of the greatest
criminals/murderers of all time!” Dr. Fudd-Einstein said in awe and sarcasm.
“I would never have suspected her, but I guess it makes sense. She is older, so she doesn’t need
near the amount of sleep as a younger person. She has access to all of the vehicles so she can copy
license plate numbers and follow them to their houses. She has the finances to pay for this kind
of massive campaign on behalf of the Mob, and she hates anyone that even remotely smells like
smoke,” Professor Bird explained, wishing to have his hit of nicotine at this time.
“Well, I guess there is nothing left to do but to call Police Chief Runner and have him arrest
her,” Dr. Fudd-Einstein explained as he began dialing. “What I can’t understand is how in the world
the Police Chief sent me all of this information and somehow seemed to screw it up.”
“What do you mean?” inquired Professor Bird.
“Well, look here. The data file from the Chief’s email shows 168 murders, but there have only
been 166. This doesn’t make any sense. I’ll have to straighten it out. Hey, wait a minute. Look at
this. Person #167 and Person #168 seem to match our stats. But how can that be?”
It was at this moment that our two heroes were shot from behind and fell over the computer,
dead. The killer hit Delete on the computer and walked out slowly (considering they had arthritis)
and cackling loudly in the now quiet computer lab.
And so, I guess my question to you the reader is, did Granny murder 168 people, or did the
murderer slip through the cracks of justice? You be the statistician and come to your own conclu-
387
sion.
Detective Pyork E. Pig
•i- •}• pj -f* •i' -i-
388
APPENDIX H. RCMDRTESTDRIVE STORY
Bibliography
[1] Daniel Adler and Duncan Murdoch, rgl: 3D visualization device system (OpenGL), 2009. R
package version 0.87. Available from: http://CRAN.R-project.org/package=rgl. 3
[2] A. Agresti and B. A. Coull. Approximate is better than "exact" for interval estimation of
binomial proportions. The American Statistician, 52:119-126, 1998.
[3] Alan Agresti. Categorical Data Analysis. Wiley, 2002. 223
[4] TomM. Apostol. Calculus, volume II. Wiley, second edition, 1967. 361
[5] TomM. Apostol. Calculus, volume I. Wiley, second edition, 1967. 361
[6] Robert B. Ash and Catherine Doleans-Dade. Probability & Measure Theory. Harcourt Aca-
demic Press, 2000. 361
[7] Peter J. Bickel and Kjell A. Doksum. Mathematical Statistics, volume I. Prentice Hall, 2001.
234
[8] Patrick Billingsley. Probability and Measure. Wiley Interscience, 1995. 123,361
[9] Ben Bolker. emdbook: Ecological models and data (book support), 2009. R package version
1.2. Available from: http://CRAN.R-project.org/package=emdbook. 180
[10] Bruce L. Bowerman, Richard O’Connell, and Anne Koehler. Forecasting, Time Series, and
Regression: An Applied Approach. South-Western College Pub, 2004.
[11] P. J. Brockwell and R. A. Davis. Time Series and Forecasting Methods. Springer, second
edition, 1991. 24
[12] Neal L. Carothers. Real Analysis. Cambridge University Press, 2000. 361
[13] George Casella and Roger L. Berger. Statistical Inference. Duxbury Press, 2002. viii, 177,
193,211
[14] Scott Chasalow. combinat: combinatorics utilities, 2009. R package version 0.0-7. Available
from: http://CRAN.R-project.org/package=combinat. 69
[15] Erhan Cinlar. Introduction to Stochastic Processes. Prentice Hall, 1975.
[16] William S. Cleveland. The Elements of Graphing Data. Hobart Press, 1994.
389
390
BIBLIOGRAPHY
[17] Fortran code by Alan Genz and R code by Adelchi Azzalini. mnormt: The multivariate
normal and t distributions, 2009. R package version 1.3-3. Available from: http://CRAN.
R-project.org/package=mnormt. 180
[18] R core members, Saikat DebRoy, Roger Bivand, and others: see COPYRIGHTS file in the
sources, foreign: Read Data Stored by Minitab, S, SAS, SPSS, Stata, Systat, dBase, ..., 2010.
R package version 0.8-39. Available from: http://CRAN.R- project. org/package=
foreign. 4
[19] Peter Dalgaard. Introductory Statistics with R. Springer, 2008. Available from: http:
//staff . pubhealth . ku . dk/~pd/ISwR . html. viii
[20] A. C. Davison and D. V. Hinkley. Bootstrap Methods and Their Applications. Cambridge
University Press, 1997.
[21] Thomas J. DiCiccio and Bradley Efron. Bootstrap confidence intervals. Statistical Science,
11:189-228,1996.
[22] Evgenia Dimitriadou, Kurt Hornik, Friedrich Leisch, David Meyer, and Andreas Weingessel.
el071: Misc Functions of the Department of Statistics (el071 ), TU Wien, 2009. R package
version 1.5-22. Available from: http://CRAN.R-project.org/package=el®71. 40
[23] Richard Durrett. Probability: Theory and Examples. Duxbury Press, 1996.
[24] Rick Durrett. Essentials of Stochastic Processes. Springer, 1999.
[25] Christophe Dutang, Vincent Goulet, and Mathieu Pigeon, actuar: An r package for actuarial
science. Journal of Statistical Software, 2008. to appear. 162
[26] Brian Everitt. An R and S-Plus Companion to Multivariate Analysis. Springer, 2007.
[27] Gerald B. Folland. Real Analysis: Modern Techniques and Their Applications. Wiley, 1999.
361
[28] John Fox. Applied Regression Analysis, Linear Models, and Related Methods. Sage, 1997.
308
[29] John Fox. An R and S Plus Companion to Applied Regression. Sage, 2002. 315
[30] John Fox. car: Companion to Applied Regression, 2009. R package version 1.2-16. Available
from: http://CRAN.R-project.org/package=car. 310
[31] John Fox, with contributions from Liviu Andronic, Michael Ash, Theophilius Boye, Stefano
Calza, Andy Chang, Philippe Grosjean, Richard Heiberger, G. Jay Kerns, Renaud Lancelot,
Matthieu Lesnoff, Uwe Ligges, Samir Messad, Martin Maechler, Robert Muenchen, Duncan
Murdoch, Erich Neuwirth, Dan Putler, Brian Ripley, Miroslav Ristic, and Peter Wolf. Rcmdr:
R Commander, 2009. R package version 1.5-4. Available from: http://CRAN.R-project.
org/package=Rcmdr. 288
[32] Michael Friendly. Visualizing Categorical Data. SAS Publishing, 2000.
BIBLIOGRAPHY
391
[33] Andrew Gelman, John B. Carlin, Hal S. Stern, and Donald B. Rubin. Bayesian Data Analysis.
CRC Press, 2004. 175
[34] Alan Genz, Frank Bretz, Tetsuhisa Miwa, Xuefei Mi, Friedrich Leisch, Fabian Scheipl, and
Torsten Hothorn. mvtnorm: Multivariate Normal and t Distributions, 2009. R package ver-
sion 0.9-8. Available from: http://CRAN.R-project.org/package=mvtnorm. 180
[35] Rob Goedman, Gabor Grothendieck, Spren Hpjsgaard, and Ayal Pinkus. Ryacas: R interface
to the yacas computer algebra system, 2008. R package version 0.2-9. Available from:
http : //ryacas . googlecode . com. 174
[36] Gregor Gorjanc. Using sweave with lyx. R News, 1:2-9,2008. 380
[37] Charles M. Grinstead and J. Laurie Snell. Introduction to Probability. American Mathemati-
cal Society, 1997. Available from: http://www.dartmouth.edu/~chance/. viii
[38] Bettina Grlin and Achim Zeileis. Automatic generation of exams in R. Journal of Statistical
Software, 29(10): 1—14. 2009. Available from: http://www.jstatsoft.org/v29/ilQ/.
viii
[39] Frank E Harrell, Jr and with contributions from many other users. Hmisc: Harrell Miscel-
laneous, 2009. R package version 3.7-0. Available from: http : //CRAN . R-proj ect . org/
package=Hmisc.
[40] Richard M. Heiberger. HH: Statistical Analysis and Data Display: Heiberger and Hol-
land, 2009. R package version 2.1-32. Available from: http://CRAN.R-project.org/
package=HH. 262
[41] Richard M. Heiberger and Burt Holland. Statistical Analysis and Data Display: An Inter-
mediate Course with Examples in S-Plus, R, and SAS. Springer, 2004. Available from:
http : //astro . temple . edu/~rmh/HH/.
[42] Richard M. Heiberger and Erich Neuwirth. R Through Excel: A Spreadsheet In-
terface for Statistics, Data Analysis, and Graphics. Springer, 2009. Available
from: http://www.springer.com/statistics/computanional-i-statistics/book/
978- 1-44 19- QQ 51- 7.
[43] Robert V. Hogg, Joseph W. McKean, and Allen T. Craig. Introduction to Mathematical Statis-
tics. Pearson Prentice Hall, 2005. 193
[44] Robert V. Hogg and Elliot A. Tanis. Probability and Statistical Inference. Pearson Prentice
Hall, 2006. viii
[45] Torsten Hothorn and Kurt Hornik. exaclRcinkTests: Exact Distributions for Rank and Permu-
tation Tests, 2006. R package version 0.8-18.
[46] Torsten Hothorn, Kurt Hornik, Mark A. van-de Wiel, and Achim Zeileis. Implementing a
class of permutation tests: The coin package. Journal of Statistical Software, 28:1-23, 2008.
[47] Norman L. Johnson, Samuel Kotz, and N. Balakrishnan. Continuous Univariate Distribu-
tions, volume 1. Wiley, second edition, 1994. 143
392
BIBLIOGRAPHY
[48] Norman L. Johnson, Samuel Kotz, and N. Balakrishnan. Continuous Univariate Distribu-
tions, volume 2. Wiley, second edition, 1995. 143
[49] Norman L. Johnson, Samuel Kotz, and N. Balakrishnan. Discrete Multivariate Distributions.
Wiley, 1997. 165
[50] Norman L. Johnson, Samuel Kotz, and Adrienne W. Kemp. Univariate Discrete Distributions.
Wiley, second edition, 1993. Ill
[51] Roger W. Johnson. How many fish are in the pond? Available from: http://www.rsscse.
org . uk/ts/gtb/ j ohnson3 . pdf.
[52] G. Jay Kerns, prob: Elementary Probability on Finite Sample Spaces, 2009. R package
version 0.9-2. Available from: http://CRAN.R-project.org/package=prob. 68
[53] G. Jay Kerns, with contributions by Theophilius Boye, Tyler Drombosky, and adapted from
the work of John Fox et al. RcmdrPlugin.IPSUR: An IPSUR Plugin for the R Commander,
2009. R package version 0.1-6. Available from: http://CRAN.R-project.org/package=
RcmdrPlugin . IPSUR.
[54] Samuel Kotz, N. Balakrishnan, and Norman L. Johnson. Continuous Multivariate Distribu-
tions, volume 1: Models and Applications. Wiley, second edition, 2000. 165, 179
[55] Max Kuhn and Steve Weaston. odfWeave: Sweave processing of Open Document Format
(ODF) files, 2009. R package version 0.7.10. 375
[56] Michael Lavine. Introduction to Statistical Thought. Lavine, Michael, 2009. Available from:
http : //www . math . umass . edu/~lavine/Book/book . html. viii
[57] Peter M. Lee. Bayesian Statistics: An Introduction. Wiley, 1997. 175
[58] E. L. Lehmann. Testing Statistical Hypotheses. Springer- Verlag, 1986. viii
[59] E. L. Lehmann and George Casella. Theory of Point Estimation. Springer, 1998. viii
[60] Uwe Ligges. Accessing the sources. RNews, 6:43-45, 2006. 10, 12
[61] Uwe Ligges and Martin Machler. Scatterplot3d - an r package for visualizing multi-
variate data. Journal of Statistical Software, 8(11): 1-20, 2003. Available from: http:
//www. jstatsoft.org. 188
[62] Jan R. Magnus and Heinz Neudecker. Matrix Differential Calculus with Applications in Statis-
tics and Econometrics. Wiley, 1999. 361
[63] John Maindonald and John Braun. Data Analysis and Graphics Using R. Cambridge Univer-
sity Press, 2003.
[64] John Maindonald and W. John Braun. DAAG: Data Analysis And Graphics data and func-
tions, 2009. R package version 1.01. Available from: http://CRAN.R-project.org/
package=DAAG.
[65] Ben Mezrich. Bringing Down the House: The Inside Story of Six M.I.T. Students Who Took
Vegas for Millions. Free Press, 2003. 80
BIBLIOGRAPHY
393
[66] Jeff Miller. Earliest known uses of some of the words of mathematics. Available from:
http : // j ef f 56® . tripod . com/mathword . htral. 21
[67] John Neter, Michael H. Kutner, Christopher J. Nachtsheim, and William Wasserman. Applied
Linear Regression Models. McGraw Hill, third edition, 1996. 222, 285, 308, 315
[68] R Development Core Team. R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria, 2009. ISBN 3-900051-07-0. Available
from: http://www.R-project.org. 356
[69] C. Radhakrishna Rao and Helge Toutenburg. Linear Models: Least Squares and Alternatives.
Springer, 1999. 285, 290
[70] Sidney I. Resnick. A Probability Path. Birkhauser, 1999. 177, 361
[71] Maria L. Rizzo. Statistical Computing with R. Chapman & Hall/CRC, 2008. 319
[72] Christian P. Robert and George Casella. Monte Carlo Statistical Methods. Springer, 2004.
319
[73] Kenneth A. Ross. Elementary Calculus: The Theory of Calculus. Springer, 1980.
[74] P. Ruckdeschel, M. Kohl, T. Stabla, and F. Camphausen. S4 classes for distributions. R News,
6(2):2-6, May 2006. Available from: http://www.uni-bayreuth.de/departments/
math/org/mathe7/DISTR/distr .pdf. 114, 118, 148, 197,214
[75] Deepayan Sarkar. lattice: Lattice Graphics, 2009. R package version 0.17-26. Available
from: http://CRAN.R-project.org/package=lattice. 286
[76] F. E. Satterthwaite. An approximate distribution of estimates of variance components. Bio-
metrics Bulletin, 2:110-114, 1946. 222
[77] Luca Scrucca. qcc: an r package for quality control charting and statistical process control. R
News, 4/1:11-17, 2004. Available from: http://CRAN.R-project.org/doc/Rnews/. 28
[78] Robert J. Serfling. Approximation Theorems of Mathematical Statistics. Wiley, 1980.
[79] Greg Snow. TeachingDemos: Demonstrations for teaching and learning, 2009. R package
version 2.5. Available from: http://CRAN.R-project.org/package=TeachingDemos.
197
[80] James Stewart. Calculus. Thomson Brooks/Cole, 2008. 361
[81] Stephen M. Stigler. The History of Statistics: The Measurement of Uncertainty before 1900.
Harvard University Press, 1986.
[82] Gilbert Strang. Linear Algebra and Its Applications. Harcourt, 1988. 361
[83] Barbara G. Tabachnick and Linda S. Fidell. Using Multivariate Statistics. Allyn and Bacon,
2006. 40, 315
394
BIBLIOGRAPHY
[84] W. N. Venables and B. D. Ripley. Modern Applied Statistics with S. Springer, New York,
fourth edition, 2002. ISBN 0-387-95457-0. Available from: http://www.stats.ox.ac.
uk/pub/MASS4. 225
[85] William N. Venables and David M. Smith. An Introduction to R, 2010. Available from:
http://www.r-project.org/Manuals. 8, 11,356
[86] John Verzani. UsingR: Data sets for the text "Using R for Introductory Statistics". R package
version 0.1-12. Available from: http://www.math.csi.cuny.edu/UsingR. 20
[87] John Verzani. Using R for Introductory Statistics. CRC Press, 2005. Available from: http:
//www. math, csi . cuny.edu/UsingR/. viii
[88] B. L. Welch. The generalization of "student’s" problem when several different population
variances are involved. Biometrika, 34:28-35, 1947. 222
[89] Hadley Wickham. Reshaping data with the reshape package. Journal of Statistical Software ,
21(12), 2007. Available from: http://www.jstatsoft.org/v21/il2/paper. 354
[90] Hadley Wickham. ggplot2: elegant graphics for data analysis. Springer New York, 2009.
Available from: http://had.co.nz/ggplot2/book. viii
[91] Graham Williams, rattle: A graphical user interface for data mining in R using GTK, 2009.
R package version 2.5.12. Available from: http://CRAN.R-project.org/package=
rattle. 6
[92] Peter Wolf and Uni Bielefeld, aplpack: Another Plot PACKage: stem.leaf bagplot, faces,
spin3R, and some slider functions, 2009. R package version 1.2.2. Available from: http:
//CRAN . R-pro j ect . org/package=aplpack. 24
[93] Achim Zeileis and Torsten Hothorn. Diagnostic checking in regression relationships. R News,
2(3):7— 10, 2002. Available from: http://CRAN.R-project.org/doc/Rnews/. 272
Index
#, 7
dot plot, see { strip chart } 1 9
. Internal, 12
DOTplot, 20
.Primitive, 12
double, 8
.RData, 15
dump, 14
.Rprofile, 15
?, 12
ecdf, 126
??, 13
Emacs, 5
[], 9
Empirical distribution, 126
ESS, 5
Active data set, 27
event, 73
apropos, 13
example, 13
as . complex, 8
exp, 7
barplot, 28
fitted values, 291
boot, 325
boot.ci, 326
hat matrix, 291
help, 12
c, 9
help. search, 13
cex. names, 28
help. start, 12
complex, 8
hist, 21
confint, 295
Histogram, 20
CRAN, 14
history, 14
Data sets
install .packages, 4
cars, 250
intersect, 10
discoveries, 19 ,
LakeHuron, 24
’
precip, 18, 21
LETTERS, 9
rivers, 18, 323
letters, 9
state . abb, 25
library, 4
state . division, 28
likelihood function, 253, 290
state .name, 25
lm, 290
state. region, 27
Is, 14
trees, 286
UKDriverDeaths, 24
maximum likelihood, 253, 290
Deducer, 6
model
depths, 24
multiple linear regression,
digits, 7
model matrix, 286
395
396
INDEX
model .matrix, 291
sample, 127
mutually exclusive, 73
sample space, 67
scan, 9
NA, 8
Sciviews-K, 5
names, 18
seq, 9
NaN, 8
sessionlnfo, 14
nominal data, 25
sigma. test, 239
normal equations, 290
sqrt, 7
stem, leaf, 24
objects, 14
str, 18, 27
options, 7
strip chart, 19
ordinal data, 25
stripchart, 19
par, 28
pareto. chart, 28
pie, 32
plot, 24
Poisson process, 135
Poor Man’s GUI, 6
power . examp, 243
predict, 261
prop. table, 27
R Commander, 6
R Editor, 5
R Graph Gallery, 14
R Graphical Manual, 14
R packages
UsingR, 20
R packages
aplpack, 24
distr, 147
qcc, 28
RcmdrPlugin . IPSUR, 28
R-Forge, 14
R-Wiki, 14
Rattle, 6
regression assumptions, 250
regression line, 250
remove, 14
replicate, 243
response vector, 286
rev, 11
Rprofile . site, 15
RSiteSearch, 13
RWinEdt, 5
t . test, 237
The R-Project, 14
Tinn-R, 5
typeof, 8
urnsamples, 70
UseMethod, 11
wilcox.test, 11
z . test, 237
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11